<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[ddouble - Medium]]></title>
        <description><![CDATA[thoughts and ideas from ddouble - Medium]]></description>
        <link>https://medium.com/ddouble?source=rss----f293534d269e---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>ddouble - Medium</title>
            <link>https://medium.com/ddouble?source=rss----f293534d269e---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 02:23:30 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/ddouble" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Ten months ago, I made a digital clone of myself. Now, she’s back.]]></title>
            <link>https://medium.com/ddouble/ten-months-ago-i-made-a-digital-clone-of-myself-now-shes-back-a9ef3538c42d?source=rss----f293534d269e---4</link>
            <guid isPermaLink="false">https://medium.com/p/a9ef3538c42d</guid>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[messaging]]></category>
            <category><![CDATA[mobile]]></category>
            <category><![CDATA[conversational-ui]]></category>
            <category><![CDATA[chatbots]]></category>
            <dc:creator><![CDATA[Sandi MacPherson]]></dc:creator>
            <pubDate>Tue, 14 Feb 2017 18:21:48 GMT</pubDate>
            <atom:updated>2017-02-14T18:21:48.437Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pmPLgI5wkDTjw3arFPwfMA.jpeg" /></figure><p>I created a simple scripted ‘mebot’ last year dubbed <a href="https://chatbotsmagazine.com/i-built-a-personal-chat-bot-why-i-love-it-and-youll-love-yours-too-6c2d019f3725">SandiMacBot</a>, that allowed you to have a conversation with a bot version of me. She explored some interesting ideas:</p><ul><li>How do we scale human communication?</li><li>Can people build on their existing relationship with you (or create a new relationship) by interacting with your ‘digital double’?</li><li>How have existing social platforms failed to capture intimate, visceral content, leading to a shift to messengers?</li><li>Why are all products clambering to further increase synchronicity?</li></ul><p>Developing solutions to these questions is the purpose behind ddouble (short for ‘digital double’), a new project I’ve been spending time on. We’ve spent the past few months narrowing in on these ideas, and are excited to release a much improved version (but still very much a prototype).</p><p>My digital double is <a href="http://m.me/1415801425117446">on Facebook’s Messenger</a>, ready to chat with my friends, acquaintances, and strangers about a variety of topics, share what I’m up to, and otherwise allow people to interact me — without my actual involvement.</p><p>This next version of my digital double can chat with anyone and everyone about:</p><ul><li>What I’ve been up to — my double might let you know about coffee shops I’ve visited, what my Friday night was like, or share a recent photo, as well as where I am or what I was up to most recently.</li><li>My favorite books on startups, graphic novels I’ve read, and my favorite books of the past year, along with articles I recommend people read on everything from product management to diversity in tech.</li><li>Music I’m listening to lately, products I’ve bought, and my favorite apps across various categories.</li><li>Many of the details of my backstory — where I’m from, what I studied in school, where I’ve lived. Random personal factoids too (SPOILER: I’m allergic to mussels!).</li><li>…and of course, share random easter eggs :D</li></ul><h3>Chat to my double on Messenger <a href="http://m.me/1415801425117446">here</a>!</h3><p><em>One quick note: The underlying infrastructure is a bit jankier (i.e. sorry in advance for delays and weird bugs — startups are all about tradeoffs!), but the overall experience is much more real, with much richer content :)</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a9ef3538c42d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ddouble/ten-months-ago-i-made-a-digital-clone-of-myself-now-shes-back-a9ef3538c42d">Ten months ago, I made a digital clone of myself. Now, she’s back.</a> was originally published in <a href="https://medium.com/ddouble">ddouble</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why did the founder of a $2.0B tech co build this bot? He wanted it for himself]]></title>
            <link>https://medium.com/ddouble/why-did-the-founder-of-a-2-0b-tech-co-build-this-bot-he-wanted-it-for-himself-dbf687ee65f0?source=rss----f293534d269e---4</link>
            <guid isPermaLink="false">https://medium.com/p/dbf687ee65f0</guid>
            <category><![CDATA[ppc-marketing]]></category>
            <category><![CDATA[hubspot]]></category>
            <category><![CDATA[bot-teardown]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[saas]]></category>
            <dc:creator><![CDATA[Sandi MacPherson]]></dc:creator>
            <pubDate>Tue, 29 Nov 2016 19:03:48 GMT</pubDate>
            <atom:updated>2016-11-29T19:03:48.020Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/867/1*aJuOIqKv5gkRDNvSAw4F9Q.png" /></figure><p>Know what information you need, but don’t want to go through the hassle of all the manual research to get it? That’s one of the core insights behind Growthbot, a new Slackbot from Hubspot. If you’re a marketing, sales, or growth professional, there’s a lot of information that you access everyday as you go about your job — and that’s exactly what Growthbot is here to help with.</p><p>I really enjoyed using Growthbot over the past couple of weeks. It fits very nicely into how I think about what ‘makes sense’ right now for conversation-based products (more on that later…) — so, on with the teardown!</p><h3>About</h3><p>Growthbot was released earlier this year, on July 14 as a Slack bot. It came directly from the hands of HubSpot Founder and CTO Dharmesh Shah:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/264/1*GeWf9bFe2bJ9kjdeGFy5cA.png" /></figure><p>…who built it for a pretty straight-forward reason:</p><blockquote>I wanted it for myself. So I built it for myself. — Dharmesh Shah</blockquote><p>But also, of course, because it makes sense for HubSpot to create a tool like this for both their current and future customers:</p><blockquote>My hope is that not only will the tool be useful to existing HubSpot customers but that it also extends HubSpot’s brand and reach to prospective customers. HubSpot believes strongly in the idea that you should create value for your market <strong>before</strong><em> </em>you to try to extract value.</blockquote><blockquote>We have reached millions [of users] through our free content, free education and training, and free tools. GrowthBot extends that idea by offering another useful (and free) tool built for our target market. — Dharmesh Shah, in <a href="http://www.inc.com/jeff-haden/confessions-of-a-business-bot-builder-the-founder-of-a-175-billion-company-who-.html">Inc. Magazine</a></blockquote><h4>What does it do?</h4><p>Growthbot responds to common questions a web-focused sales/marketing/growth person may have. For example:</p><ul><li>What keywords is website.com buying?</li><li>How many monthly visits does website.com get?</li><li>What is the top performing content on Specific Topic?</li><li>What marketing software does website.com use?</li></ul><p>There are also some general company queries that Growthbot can respond to, such as:</p><ul><li>What is the stock price of $ticker?</li><li>How many employees does Company have?</li><li>How much funding has Company raised?</li></ul><h4>Talking to Growthbot</h4><p>To get started, you have to follow the standard ‘Add to Slack’ process, starting at the <a href="https://growthbot.org/">Growthbot website</a>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/784/1*kLHNuQ2NAo426Jn1sM-3Bg.png" /></figure><p>Once you’ve added Growthbot, there’s the option to add it to other channels, to move the bot from outside of your private DM-based conversation. Growthbot sits in your DM panel, with the inactive indicator. There seems to be no best practice yet around what type of status bots should be in — while it’s true that they’re not actually ‘active’, it’s also true that their activity state is solely determined by the user, meaning they’re basically active all the time (or at least when you need them to be).</p><h4>Natural Language Processing and tone</h4><p>There’s some NLP being used to understand the questions that someone asks to Growthbot. It’s pretty good, but I definitely hit some errors when some of my questions were not asked in a specific way:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/476/1*Ct_Bub_YtjdkQrCk153_gA.png" /></figure><p>That said, if you have an understanding of specifically what you’re looking for, simple keyword queries without any formal sentence structure will return what you’re looking for:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/435/1*qnu-5gN6A6Mru3mZB3Gusw.png" /></figure><p>I’m assuming that once the HubSpot team has some more data around the types of questions people are asking, what returned errors, and what the intention of those questions were, they’ll be able to improve the accuracy of responses and lower the rate of returning an error response.</p><p>Growthbot is also pretty humble and polite. If you hit an error, Growthbot is self-effacing, and almost embarrassed that the answer can’t be found:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/423/1*AsgosvpLtag6V7v31Z_85A.png" /></figure><p>There’s another type of error too, that reinforces the idea that Growthbot is pretty great, and is only going to get better — so stick around:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/476/1*Ct_Bub_YtjdkQrCk153_gA.png" /></figure><p>There’s no need to end your queries with a question mark, as Growthbot assumes that anything you send as a message is a question — a fair assumption for the product and one that save key strokes, too.</p><h4>Behind the scenes</h4><p>Growthbot responds to all of these requests by cobbling together various APIs — some from Hubspot, some from other services that generate intelligence on web-based products (e.g. Google Analytics, Alexa, SimilarWeb). As you can see in the previous screenshot, the source of the query is given at the bottom of the response from Growthbot if you want to dive deeper into the information provided.</p><h3>The Growthbot experience</h3><p>So — what happens when you start asking Growthbot questions? What types of responses can you expect? Let’s get into what it’s like to interact with Growthbot…</p><h4>Company information</h4><p>If there’s a particular company that you’re interested in, Growthbot can respond to questions about company’s operations, people, and market information. For example, you can ask Growthbot to find the Founder of a company:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/440/1*wZ6-hxu4xawbfixga7yB5A.png" /></figure><p>… list the competitors to any website:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/420/1*qWx_r9zjCCbXDOR8r7GBag.png" /></figure><p>…or show details about the company, including where their (head) office is located, how many employees they have, and all of their online profiles:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*bhXDGyMUDgh7iABCxPnx1w.png" /></figure><h4>Marketing tools and behaviors</h4><p>One of the most compelling parts of Growthbot is how quick and easy it is to find information about the marketing tactics and tools of any website. IMHO, this is where the product really shines.</p><p>Growthbot can tell you what keywords a website is ranking for:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/489/1*3eDXb5neRdAQd38snYz18A.png" /></figure><p>…what keywords they’re buying, and for how much:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/495/1*e7AKrdwx3ZzkajM5Ie_B-g.png" /></figure><p>…how much of their web traffic is organic, and what is the value of that traffic:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/487/1*RP1iU9zjLO0WQQFee9TrAQ.png" /></figure><p>…how many total monthly visits their website gets:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/477/1*E93kQxtvI8-PiQzlEfC4MA.png" /></figure><p>… and lots more.</p><p>I’m uncertain as to how accurate all of these responses are, and I haven’t been able to dive into the methods that HubSpot or the supporting services are using to provide the information. For example, as the Founder of Quibb, I know that this list of marketing services isn’t accurate (some are missing):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/477/1*h7cuhT2Y9D89WLtomHwGpw.png" /></figure><p>At the same time, if you’re a marketing/sales/growth professional and Growthbot is simply a tool to help streamline your process for finding the information you need and already use to do your job, you’re hopefully already familiar with the potential errors these services might let slip through, and the process they go through to serve you that info.</p><h3>Errors and what I’m hoping for next</h3><p>There was one error that I hit fairly often. Unless you append a ‘.com’ to a company or website you’re asking questions about, Growthbot will assume you’re looking for a teammate on Slack with that username:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/477/1*G1rOGyrzLtPIBsjSGrjRaQ.png" /></figure><p>I’d assume that people using the tool would be more likely to use Growthbot to find information about companies than their colleagues, so I’m unsure why this product decision was made ¯\_(ツ)_/¯</p><h4>Can I trust the answer?</h4><p>I was served some wrong information when asking questions that I expected to be in-scope:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/477/1*SLQ7Cn4FgDIdN-omB5jlgA.png" /></figure><p>This is one of my main problems overall with Growthbot, and also one of the more difficult problems the product has to solve. If you’re browsing a website, and actually clicking the buttons, reading the menus, glancing at the heading of the table — you know whether or not the information you’re being shown is the information you’re looking for. This isn’t the case with Growthbot.</p><p>There were also some other (imho) more obvious, clearly articulated questions that gave me incorrect answers:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/472/1*5KjpcGxtTEDLfzoiZXz0Yw.png" /></figure><p>On several occasions with Growthbot, I was unsure — is this actually the right answer to the question I asked? I have an understanding of what I’m looking for when I search for ‘Top post on Poodles’, but what exactly does this list mean?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*LDj6l5TDHNiANGRhfwlz0g.png" /></figure><p>Are these the top posts on poodles from the start of time? In English? Is it based only on total number of page views, or shares, or some combination of those and other metrics?</p><p>While listing the source is nice, if anything I would appreciate having some sort of a glossary, potentially listed on the Growthbot website. This document could list exactly where the information is coming from, how it’s calculated, etc. If I’m going to be using this information for anything more than navel-gazing or quick gut checks, I need to make sure I have an understanding on accuracy, source, time horizon it was pulled from, etc etc.</p><h4>HubSpot-only features</h4><p>I’m not a HubSpot customer, so there were a few features that I was unable to use. For example, for any company you ask about via Growthbot you can ‘add to crm’, or ‘add to watchlist’. As a non-HubSpot user, these commands (shown in the UI as buttons) were a bit confusing. Clicking on them didn’t do anything, except display a simple update note.</p><p>I’m assuming (though I’m not positive, and am unsure why) the ‘show employees’ button is also a HubSpot-only feature, as that button didn’t yield any output for me:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/457/1*EqPfCNo384cSsjbiMXedHg.png" /></figure><p>…or perhaps it was just an error?</p><p>This points to another issue I had with Growthbot — some interactions that I made where I expected an output yielded nothing, leaving me with a confused feeling and wondering what I did wrong. Not a good position to put the user in!</p><h4>Limits of a list-based output</h4><p>One final tweak that I’d like to see would be how Growthbot displays list answers that are &gt;10 entries. For example, almost 40k (!) companies in San Francisco use MailChimp… but Growthbot’s output is a 10 item list, of unknown order (size of company? alphabetical? random?):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/404/1*-89qBbghcijG-Hzw7LiiPA.png" /></figure><p>Here, the ‘enter its number’ action doesn’t really allow me to do much. And there’s no other way to expand beyond that list of ten.</p><p>Similarly, while Growthbot is pretty new and cutting-edge… emoji aren’t quite understood, yet 😆</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/390/1*Ao4M8hIe5aBtucJtnNr3RQ.png" /></figure><p>But, what is really nice with that flow — if you type ‘2’ after the error, Growthbot remembers the question that you asked previously, and what that new entry relates to:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/505/1*4zF-r06Y9EnRTtFB7XiBjQ.png" /></figure><h3>🎉🎉🎉</h3><h4>What about mobile?</h4><p>One of the biggest pieces missing from Growthbot are metrics and information on mobile-based companies and products. I assume that Hubspot is used almost exclusively for heavily web-focused companies.</p><p>Still, it was a disappointing that Growthbot couldn’t give me any information about mobile apps (and instead, I think, gave me content about it?):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/486/1*HArSSQaYISVaGCfUZBMDRg.png" /></figure><h3>Why I ❤️ Growthbot</h3><p>When a sales/marketing/growth professionals needs to access some information to do their job, their process looks something like this:</p><ul><li>encounter a question</li><li>think about/ask colleague/research where to find the information needed</li><li>navigate to that website</li><li>look around, confirm how the UI works</li><li>interact with the website (probably via search field) to query for the information you need</li><li>website serves response</li><li>comprehend UI to extract core content (is it a graph? table? text?)</li></ul><p>Some of these steps can be cut out over time (i.e. the two points around comprehending how the UI of the website works or displays information, where to find the info you need, etc.), but otherwise this is the typical process we all go through.</p><p>Let’s compare that to Growthbot:</p><ul><li>encounter a question</li><li>navigate to Growthbot channel in Slack</li><li>ask Growthbot for the information you need</li><li>Growthbot serves response and core content is extracted</li></ul><p>That’s a simplified comparison, of course. It also assumes that a conversational UI creates a more seamless and accessible interaction model (which we all think is true, but are we <strong>really</strong> 100% certain? ¯\_(ツ)_/¯).</p><p>This comparison highlights what it is that Growthbot does so well, and why it’s the perfect bot — it reduces the cognitive friction for an action where the product itself is irrelevant. Getting the content or information itself to the user is all that matters, and Growthbot constructs the fewest possible barriers between that initial user intent and the content they seek.</p><p>What I’m most excited about for the future of Growthbot is the potential to string a much more complicated series of questions together, and have Growthbot understand exactly the outcome and final information I’m looking for. For example, if I want to know which companies are my competitors and which keywords all of them are buying — I can do that that now with Growthbot, sure, but it’s a lot of steps and back-and-forth. It’s cumbersome:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/519/1*ygChtpL-VHWO9GFDfREoUQ.png" /></figure><p>But it feels like Growthbot is on the path to making exactly this type of repetitive, clearly defined type of question and data collection process much simpler and faster than the current human-powered version. Just like Dharmesh — that’s a tool that I’d like to have, too.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dbf687ee65f0" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ddouble/why-did-the-founder-of-a-2-0b-tech-co-build-this-bot-he-wanted-it-for-himself-dbf687ee65f0">Why did the founder of a $2.0B tech co build this bot? He wanted it for himself</a> was originally published in <a href="https://medium.com/ddouble">ddouble</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How Twitter bots paved the way for today’s chatbots]]></title>
            <link>https://medium.com/ddouble/how-twitter-bots-paved-the-way-for-todays-chatbots-2fa103020d3e?source=rss----f293534d269e---4</link>
            <guid isPermaLink="false">https://medium.com/p/2fa103020d3e</guid>
            <category><![CDATA[messaging]]></category>
            <category><![CDATA[bot-teardown]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[twitter]]></category>
            <dc:creator><![CDATA[Sandi MacPherson]]></dc:creator>
            <pubDate>Tue, 15 Nov 2016 18:42:12 GMT</pubDate>
            <atom:updated>2016-11-15T18:45:31.702Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/824/1*aXX2N3bYeil_iPUEmMm1AQ.jpeg" /></figure><p>This week, we’re taking a bit of a detour — more like a rewind — in the Tuesday Bot Teardown series. I was chatting with <a href="https://twitter.com/fourtonfish">Stefan Bohacek</a> (of Botwiki) about the teardown series, and he suggested that I look at Twitter bots. While most Twitter bots are too simplistic to warrant an entire post, I did want to spend some time learning about the current state of Twitter bots, where they came from, their impact on peoples’ understanding of bots, and compare Twitter bots to other bots-of-the-moment that exist within a conversational UI.</p><p>Also… November is <a href="http://nabomamo.botally.net/">#NaBoMaM</a>o - A month-long celebration of non-commercial botmaking!!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/498/1*j0LNQopY8trlpeUoPlymMg.gif" /></figure><p><em>Quick note — as this is more of a research mashed with an opinion piece, I would love to hear of any other resources, or your thoughts on the importance of Twitter bots in the work we’re doing today on chatbots :)</em></p><h3>The original user-facing bots</h3><p>Today, the platforms are all working hard to convince developers to build bots for their users. Facebook’s Messenger, Slack, Kik, Telegram… they’re all fairly new to the bot game in comparison to Twitter. Just over two years ago, Twitter had 23 million active bots on their platform. And that was all without any concerted effort or attempts to entice their creation. Bravo!</p><p>Twitter bots are very different from the current wave of bots that are taking over the headlines of the Bay Area. Where most of the bots being built today have clear, explicit functions that create value for the end user in obvious ways (i.e. scheduling meetings, shopping assistants, news analysis, scaling repeatable tasks), Twitter bots have by and large existed in the realm of ‘just for fun’. Those that do provide access to information (which we’ll get into shortly) are often simply exposing information in specific ways, and not actually creating an experience. Twitter bots are subservient, quietly processing information, only to be collected or reviewed when someone decides to take action.</p><h4>But… Twitter bots can’t chat</h4><p>In today’s bot trend, we’re often coupling the idea of bots with the idea of ‘conversational UI’. Many people define bots as those that have the ability to chat with users. This is one of the biggest ways in which most Twitter bots are different. There’s often no exchange between the followers (i.e. users) of a Twitter bot and the bot itself.</p><p>There are many other differences too — there’s no onboarding, you can’t invite others to join, there’s no group experience, no education on how to use or engage with the bot, etc. They are similar in that they’re an automated, sometimes AI-driven entity that creates mostly text-based content that’s shared on a social message-based platform.</p><h4>User education and exposure</h4><p>I believe that having bots intermingled with real people on Twitter has probably been a huge boost and advantage for anyone working on bots now. The idea of a bot has been established in the minds of users, often as a funny or artistic entertainment object. You could also push this a bit further, to argue that Twitter bots also pushed forward the idea of bots existing in a conversational space. While Twitter isn’t quite the messenger-derived conversational UI of most bots today, it is true that Twitter existed for many years as an almost public texting equivalent. We can assume that seeing bots intermingled in that space made the jump of them to full one-on-one conversational UI a smoother transition.</p><h3>Learning about Twitter bots</h3><p>My process for this teardown was a bit different than previous posts. I followed just over 100 Twitter bots for 2w, and made notes throughout that time on my experience — both as they’d surface in my feed, and by browsing a list I made that included each of the bots. You can check out <a href="https://twitter.com/sandimac/lists/twitter-bots">my list here</a>. Over the past 2w, I’ve tried to note and identify what I liked or disliked, my experience as a follower of the bot, and also gave some thought to the intention of the creator of the Twitter bot.</p><p>Personally, my (purposeful) experience with Twitter bots is only a few years old, when I followed <a href="http://www.slate.com/blogs/browbeat/2012/01/20/stealth_mountain_the_twitter_bot_devoted_to_a_single_grammatical_error.html">Stealth Mountain</a> in 2012. I’ve since subscribed to some other Twitter bots over the past few years, but wanted to experience a wider array and try to understand the various types of Twitter bots that exist. Otherwise, I have a twitter bot ‘running’ the <a href="https://twitter.com/Quibb">Twitter account</a> of my company Quibb —it tweets out a link to the most popular article shared on Quibb.</p><h4>Making a Twitter bot</h4><p>I think one of the reasons why there are so many Twitter bots is that they’re very simple to build. The Twitter API is easy to work with, and the approval process and permissions aren’t nearly as restrictive as other platforms. Discoverability on Twitter is also much more straight-forward than other platforms — if gaining an audience is your goal.</p><p>It also basically runs itself. There’s very little upfront work or platform ‘maintenance’ that you have to worry about — there’s no need to go back and approve or accept friends or connections, there’s no need to link to a privacy policy or terms of services page (and keep them up-to-date), etc.</p><p><em>If you’re curious to learn more about how to make a Twitter bot, Botwiki has </em><a href="https://botwiki.org/tutorials/twitterbots/"><em>some great resources</em></a><em>.</em></p><h3>Twitter bot ‘design’</h3><p>Designing a Twitter bot is very different from designing a conversational UI-based bot/chatbot that many of us are working on today. With Twitter, there are many variables that are constrained. The most obvious is exposure and therefore engagement. A Twitter bot is simply slotted into the firehose that is Twitter along with every other account that a follower follows. There’s no ability for the bot to obtain privileged placement, or to ensure visibility to their followers.</p><p>There are also the content constraints, however this is evolving and more types are becoming accessible over time. Twitter also has a very simplistic network model, with those you follow and followers. Creating groups or segmenting the bot’s network isn’t really an option.</p><h3>Twitter bots I hung out with</h3><p>In having 100 Twitter bots in my feed, I started to notice that they seem to fall into different categories. Many of those categories are based on the format of the content of the bot, while others are different due to their intention. I classified the bots into these different categories, and did a bit of a deeper dive on my favorite(s) from each.</p><p>I want to note that <a href="https://twitter.com/tullyhansen">Tully Hansen</a> wrote a <a href="https://gist.github.com/tullyhansen/7621632">really great classification system</a> for Twitter bots. He focused more on the inputs that drive the bot, whereas my (super simple, haphazard) clusters of categories below are more from the user perspective and content format.</p><p>The main two branches of the tree that he proposes is Dependent (i.e. triggers and input derived from Twitter itself), vs. Independent. For all of the bots that I spent time following, they’d all fall under the Independent category that Tully defines:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jSa9sXpdEFSMxAiwbUcd1Q.png" /></figure><h4>Images</h4><p>There are a bunch of Twitter bots that simply tweet out an image. However, how these images are generated and the complexity in creating the image varies wildly.</p><p>A few other Twitter bots in this category:</p><ul><li><a href="https://twitter.com/__birds__">Birds on wires</a>: Created by <a href="https://twitter.com/RobertVinluan">Robert Vinluan</a>, tweets an image of birds sitting on telephone/electricity wires once every 3 hours.</li><li><a href="https://twitter.com/BrutalistEngine">The Brutalist Engine</a>: Created by <a href="http://nicolastilly.fr/">Nicolas Tilly</a>, tweets out black and white images of brutalist-inspired architecture.</li><li><a href="https://twitter.com/_blackobs">Black Obs</a>: Created by <a href="http://www.matthiasplanitzer.de/black-obs/">Matthias Planitzer</a>, tweets out security camera stills from around the world in real-time.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/584/1*1QYNJl_OYXXjd759MAc9Ng.png" /></figure><ul><li><a href="https://twitter.com/WizardGenerator">Wizard Generator</a>: Created by <a href="https://twitter.com/TheWAAnderson">William Anderson</a>, I’m not exactly sure how this bot works. But the output is a very convincing and often hilarious random wizard, popping up in your feed once every 3 hours.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/582/1*Ocuo_cEpLxFEbjjUFSZwyA.png" /></figure><h3>Emoji</h3><p>While there are a few great emoji-based bots, this category of bots (which seems to be the newest) is also my least favorite. There are a few great ones (listed below first), but there seems to be many that were simply taking advantage of the recent uptick in emoji popularity, versus actually making something fun or interesting (i.e. the #5aday bot, see below).</p><p>My favorite emoji-based Twitter bot is called <a href="https://twitter.com/choochoobot">trains botting</a>, written by <a href="https://parkerhiggins.net/2016/03/new-bot-choochoobot/">Parker Higgins</a>. It creates a cute emoji-based image of a train traversing different landscapes.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/585/1*aoSJ81tN8FciF92_tHHArQ.png" /></figure><p>A few other emoji-based Twitter bots include:</p><ul><li><a href="https://twitter.com/tiny_star_field">tiny star fields</a>: Created by <a href="http://katierosepipkin.com/">Katie Rose Pipkin</a>, which randomly adds a star versus a space to create a unique sky of stars.</li><li><a href="https://twitter.com/thetinygallery">Tiny Gallery</a>: Created by <a href="https://emmawinston.me/">Emma Winston</a>, builds a small emoji room with displayed art and a handful of visitors.</li><li><a href="https://twitter.com/5adaybot">#5aday</a>: Randomly tweets 5 fruit and/or vegetable emoji every hour.</li></ul><h4>Static lists</h4><p>Listing out information that is static or unchanging is a Twitter bot category that seems pretty well established. These bots are fun because they allow you to slowly peer inside of some collection of information that, on it’s own, would be too long and boring to spend time on. Some in this category I like include:</p><ul><li><a href="https://twitter.com/everysimile">Every Simile</a>: Created by <a href="https://twitter.com/rainshapes">Tobi Hahn</a>, this bot tweets out every simile in the English language once every 15min. I’m unsure whether it will loop once it reaches the end, or if the bot will ‘die’ once all similes have been tweeted.</li><li><a href="https://twitter.com/pls_take_breaks">take a break pls</a>: Created by <a href="https://twitter.com/swayandsea">@swayandsea</a>, this simple but powerful bot tweets out an encouraging messaging along with a nudge to ‘take a break!’ once every three hours — always starting and ending with a ❤</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/584/1*ruFj325TjkUEVZzfkHldbg.png" /></figure><ul><li><a href="https://twitter.com/everyailment">Every Ailment</a>: Created by <a href="https://twitter.com/ddbeck">Daniel D. Beck</a>, this bot tweets out 2 ailments per hour (code, plus a description) for all of the medical ailments listed in the ICD-10-CM (The International Classification of Diseases, Tenth Revision).</li></ul><p>By far my favorite in this category is by far <a href="https://twitter.com/censusAmericans">censusAmericans</a>, created by <a href="https://twitter.com/jiazhang">Jia Zhang</a> of <a href="http://fivethirtyeight.com/datalab/introducing-censusamericans-a-twitter-bot-for-america/">FiveThirtyEight</a>. The Twitter bot pulls from data collected from Americans by the U. S Census Bureau from 2009 to 2013, creates mini ‘biographies’ for each person, tweeting one per hour.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/584/1*5r_dsHmJELdPil4H4NB3_g.png" /></figure><h4>Monitoring change</h4><p>This category is the one that I think has the most potential to actually be powerful. Computers are great at monitoring systems, and notifying us humans of when there’s a change. In contrast, this is something that humans are not very good at.</p><ul><li><a href="https://twitter.com/mass_shoot_bot">Mass shootings bot</a>: As the name makes clear, this bot tweets out the details of a mass shooting whenever one occurs in the US.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/584/1*oTd8-WoaJgpdzjyo2tSkaA.png" /></figure><ul><li><a href="https://twitter.com/sunrisebot">sunrisebot</a>: As the bots creator <a href="https://twitter.com/richaaaa">Richa Agarwal</a> wrote in the description for this bot “the sun is always rising, somewhere.” Approximately once every hour, the bot tweets out the details of where the sun is currently rising.</li><li><a href="https://twitter.com/nyt_diff">Editing NewYorkTimes</a>: Created by Juan E. D., this bot creates a quick visual summary of changes that are made to any NYTimes headline or abstract.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/583/1*BWgScSeQQ9531i3QvZbmGQ.png" /></figure><h4>AI (or more intense algorithmically powered) bots</h4><p>These are some really fun Twitter bots, that mash together some idea with either a fairly complicated algorithm, or much more complex AI.</p><ul><li>Glossatory: Created by <a href="https://twitter.com/bombinans">Mike Lynch</a>, this bot has much more ‘under the hood’ than most. It <a href="http://bots.mikelynch.org/glossatory/">generates</a> definitions for words, generated by a Recurrent Neural Network that was trained off a data set of &gt;80k ‘real’ definitions.</li></ul><p>My favorite in this category is <a href="https://twitter.com/unchartedatlas">Uncharted Atlas</a>, created by Glaciologist <a href="https://twitter.com/mewo2">Martin O’Leary</a>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/583/1*b_8ykD9CLH-OXCyoSWAQeA.png" /></figure><p>Martin describes how the Twitter bot runs and creates the images on <a href="http://mewo2.com/notes/terrain/">his blog</a>. What I find most compelling about this Twitter bot is the thoughtfulness Martin put into creating it. It’s based off of a ton of information on how terrain is formed, combined with a <a href="http://mewo2.com/notes/naming-language/">completely separate algorithm</a> that generates place names. The end result is a seemingly real map, complete with all of the features you’d expect on a map of a real place.</p><h3>What’s next for Twitter bots?</h3><p>In a world where more and more information becomes accessible, I can only imagine that there’s the potential for more and more Twitter bots to be built, and to actually provide input into decisions that drive people’s lives. More recently, we saw how Twitter bots can be <a href="https://www.technologyreview.com/s/602817/how-the-bot-y-politic-influenced-this-election/">extremely powerful</a> (i.e. the US presidential election) by injecting themselves into the daily news and social interactions of people’s lives.</p><p>Twitter has also recently <a href="https://dev.twitter.com/build/automation">opened up more options</a> for developers looking to build bots into both the Timeline and the Direct Message feature. Unfortunately, they seem more business-focused than in-line with the more creative, information sharing/access history of Twitter bots to-date. I’m excited to watch how Twitter bots continue to evolve, especially now as more time and thought in general is being put into the format by many more people.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2fa103020d3e" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ddouble/how-twitter-bots-paved-the-way-for-todays-chatbots-2fa103020d3e">How Twitter bots paved the way for today’s chatbots</a> was originally published in <a href="https://medium.com/ddouble">ddouble</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Is the election confusing you? This little bot will clear things up]]></title>
            <link>https://medium.com/ddouble/is-the-election-confusing-you-this-little-bot-will-clear-things-up-db25b53866df?source=rss----f293534d269e---4</link>
            <guid isPermaLink="false">https://medium.com/p/db25b53866df</guid>
            <category><![CDATA[news]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[bot-teardown]]></category>
            <category><![CDATA[messaging]]></category>
            <category><![CDATA[politics]]></category>
            <dc:creator><![CDATA[Sandi MacPherson]]></dc:creator>
            <pubDate>Tue, 25 Oct 2016 17:09:50 GMT</pubDate>
            <atom:updated>2016-10-25T17:12:15.376Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/517/1*faf8EaYZekYJs07rqpZgVg.png" /></figure><p>As a Canadian in the valley, I’ve been trying to keep up with the current election, while also learning about the US political system. I’ll read random articles friends recommend, I’ve watched the debates, and enjoy chatting about the two presidential candidates and their campaigns. As I navigate all of this, I’ve also had a bit of help recently from a Facebook Messenger bot called <a href="https://getpurple.io/">Purple</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/538/1*8f8hNCCIPrRqG1u1srKCzA.png" /></figure><p>The team (<a href="https://twitter.com/rebeccajharris">Rebecca</a> and <a href="https://twitter.com/david_heimann">David</a>) of Purple describe why they created the product:</p><blockquote>We built Purple to make staying informed ridiculously easy by taking the overwhelming stream of world news and refining it into clear, simple updates. You can always dig deeper on an issue, and you’ll always know why it matters. Think of us as your pocket-sized nerds, ready to drop knowledge whenever you need it.</blockquote><p>So… let’s dive into what works, what doesn’t, what makes Purple special, and some ideas on how to make Purple even more awesome.</p><h3>Purple on Facebook</h3><p>Purple lives on the Messenger platform at <a href="http://m.me/getprpl/">m.me/getprpl/</a>. I don’t remember how I first came across Purple, but I at the time I was trying out a few different news bots (i.e. Digg, Quartz), and wanted to see if Purple fit into what I was looking for.</p><p>Purple started off as a politics-focused newsletter in mid-2015, with the Messenger bot being released early 2016. One of Purple’s cofounders, Rebecca Harris, takes charge of collecting interesting political news every day, collating into compelling stories and areas of focus, which are then served daily by the bot. The team includes links from various sources, funny gifs, diagrams, and creates their own ‘quotes as images’ content to focus on important things being said in the world of politics.</p><p>In announcing the Messenger bot, Rebecca noted on <a href="https://www.producthunt.com/tech/a-new-media-news-bot">Product Hunt</a> that they are:</p><blockquote>“…trying to take away the pain point of weeding through the noise and overflow of content out there for you.”</blockquote><p>It’s a noble goal, one that other news products (and bots) have tried to solve. It feels most similar to <a href="https://en.wikipedia.org/wiki/Circa_News">Circa</a>, which shut down at about the same time Purple was starting up, in mid-2015. Circa also offered short summaries of content, but with a much broader range of topics. Purple’s focus at the moment is only on political news and content, mostly focused on the US and from a US-centric perspective.</p><h4>Getting Started</h4><p>The welcome from Purple is pretty fun, and the personality of the product becomes evident from the very first interaction — Purple called me ‘homie’!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*BLJc69nXKYipiEqMT7lehg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/537/1*2rpSMAnO6IbCyB7mes9qvw.png" /></figure><p>I was a little bit confused as to why the bot starts our conversation by asking me for a username. I can appreciate it with respect to trying to establish somewhat of a bond or personal connection… but that logic doesn’t quite carry over to asking me for a username versus my real name, as it seems a bit out of place in relation to a chat-based interface. And one thing to note several weeks later is that there hasn’t been another interaction where I’ve seen my username in any of our communications.</p><p>The mention of ‘message us anytime’ and ‘we’re real people’ is a nice touch, and is part of what makes Purple special (…more on that later). However, at this point I wasn’t quite sure exactly what that meant, or how I was expected to behave differently. I was also sort of ‘in the zone’ at that moment, as I had added the bot for a purpose — to get political news. I wasn’t quite interested or ready to accept the added thoughts around ‘how do i use this, beyond just pushing the buttons it displays? what does it mean that there’s ‘real people’ on the other side? what kind of questions do i ask these people’. And so, I ended up just continuing along to get some news.</p><h3>Purple’s everyday</h3><p>A typical interaction with Purple starts with a push notification, once new content for the day has been added:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/435/1*u_lZPztdHk4iMr8RH3qv0w.png" /></figure><p>Once you tap into a push notification, you can choose to either type the keyword if it’s interesting to you (i.e. the word ‘EARLY’ in this example), or tap the Lightbulb emoji:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/436/1*prA7Se51NQJLXtZALCGpSA.png" /></figure><h4>Interacting via keywords</h4><p>If you choose to type in the keyword that’s suggested, Purple will lead you through the story, highlighting what aspects it believes are most interesting and relevant. Most of the messages are no more than 10 or 12 lines long, which takes up approximately half of the total size of the chat window. Sometimes, if another piece of media is included in one of the messages, the message may expand to the size of the whole window.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/537/1*rQp3WwrimEuQqvPtn7xpMQ.png" /></figure><p>To progress through the story, the bottom of each message has a More button, which prompts the next piece of the story to appear in a message:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*57TYWVHaUqz46pJgrsU_OQ.png" /></figure><h4>Interacting via the carousel</h4><p>Taking the carousel route by tapping on the Explore button on any of the items led to a very similar type of interaction:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/537/1*Lp63PTVh6JDz9FqyoV-7Ig.png" /></figure><p>Continuing to tap the More button leads you through the story, via a series of short messages from Purple.</p><h3>How Purple chats</h3><h4>Language and tone</h4><p>Purple is clearly targeted at a younger demographic. It’s noticeable from the very first message, greeting the new user as ‘homie’. In the news summaries, they also use a lot of slang and words/phrases that are more common among young people (i.e. ‘Eh idk.’, ‘…down with that?’, ‘Yeesh.’).</p><h4>More than text messages</h4><p>Purple sends a lot of different types of messages, taking full advantage of what the Messenger platform is currently offering developers. I received lots of images and gifs from Purple, including cute images the team created to highlight important quotes that were said related to the topic at-hand:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*kk1K5XB5icIo4IkE2P9i_g.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/537/1*4chtmswOyHvUfL973f0Czw.png" /></figure><p>Some of these images were more useful than others. For example, when Purple‘s push of the day was about the current breakdown of election polls, they included a great map showing how each of the states is expected to vote:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/539/1*doOmWVEGDo7Ef-SMEa2uBg.png" /></figure><h3>Beyond the messages</h3><p>One feature that I (unfortunately) didn’t get to fully experience with Purple is the various interactive events and polls they do with their users. For example, the team brought in a special guest to do a live chat on Purple:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*8bLW_RRd9A1htNu9YxPqiw.png" /></figure><p>The team is also very open to hearing directly from their users, asking them to highlight the types of things they want Purple to cover:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*ofJ61wqii-SGK8niJ46Z9g.png" /></figure><p>These types of interactions feel really great, and engendered a sense of shared purpose as I was using the product. It also added a social level to the product that otherwise is a very one-to-one experience. That type of feature isn’t explicitly enabled via Messenger, but it’s something that the team was able to imagine and create within the restraints of the platform, which I really appreciate. They understand that reading about politics can be fun, but oftentimes talking about politics is even better, especially if you can have some ownership and help drive the direction of the conversation.</p><h4>Hi Rebecca and David!</h4><p>One of the features of Purple that makes it compelling is how founders Rebecca and David exist in relation to the product. From what I’ve gathered online, it seems like Rebecca collects and collates the information for the summaries. However, both of them are actively responding to messages that people send to Purple when/if the bot isn’t able to answer on it’s own.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/400/1*_F4gZVL9l1PBvaJd1Nukag.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/240/1*5HWcdGuNQR1S7y5YOC9feA.jpeg" /></figure><p><a href="https://en.wikipedia.org/wiki/Human-in-the-loop">Human-in-the-Loop</a> is one subset of AI that is pretty common these days, with respect to bot products. It’s where the bot itself manages as many of the interactions as it can, and then completes a ‘hand off’ to a real human if needed. This model of chatting with Rebecca and/or David feels similar to that type of interaction that’s common in other products. However, in the case of Purple, I found it felt like too much of a jump.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/437/1*8j6_3VLv8sL_efi_s3rVYA.png" /></figure><p>Interacting with Purple feels very much like interacting with a bot. When switching to chatting with Rebecca or David (indicated by a crown emoji and their name), it felt very much like conversing with a real person. The mental jump was a bit jarring — do I say thank you to Rebecca and David? If our interactions are in the same chat as the on-going dialogue with the bot, how does one mode of conversation versus another evolve? One aspect that helped was that the personality of Purple was very obvious and warm, pushing those interactions up my personal scale of human-ness. My assumption is that this then made the jump to a real human a bit less jarring than it could have been, if Purple were very cold or otherwise unhuman-like.</p><p>Mixing bots with real humans will become more common I’m sure, yet the design challenges of how to do this well haven’t yet been codified into best practices. I’m excited to see how this evolves in the coming months.</p><h3>Interacting with Purple</h3><h4>Time of day</h4><p>This is one of my biggest problems with Purple. I’ve found, in my work on Quibb, that people tend to have fairly strict habits around when they consume news. Currently, the daily push that I receive from Purple arrives… whenever. There’s no set time, and I’ve found myself looking at day-old news a couple of times.</p><h4>The content</h4><p>The actual text-based messages from Purple are fantastic. The summaries are very clearly written, and explain the current issues in a way that makes them seem very accessible — but without giving up any of the importance or weight that many of them carry. And there’s typically links to supporting information if you’re curious to learn more. I found the length to be appropriate as well. None of the stories went into too much details, leaving me with a sense of ‘omg please make it stop’. If anything, some of them probably could have been a bit longer.</p><p>However, knowing that the content is currently being written by one person on the small 2-person team, it seems unlikely that the product will be able to deliver more human-created content anytime in the near future. I do hope that investing in this part of the company is something that the Purple teams spends time on, for my own sake!</p><h4>Emoji can be 🌭 sometimes</h4><p>While I’m a big emoji fan… they’re not always the clearest way to communicate.</p><p>Each day, Purple sends a short summary and 2 buttons to take my next action:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/537/1*DGnneu1XV7JQIg1alW7QYg.png" /></figure><p>The choice of a Lightbulb and Question mark were never quite clear to me. I assumed at first that the Question mark was to indicate ‘i’m interested, i have questions on this topic! i want to learn more!’, but it actually leads to the Help module. Similarly, I found the Lightbulb emoji to be equally confusing, considering the text in the message uses the keyword ‘highLIGHTS’, which at first made me think that the lightbulb was a shortcut to the Highlights. Ugh.</p><p>I had further confusion around the emoji that are presented at the end of each summary message.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*57TYWVHaUqz46pJgrsU_OQ.png" /></figure><p>Would the Lightbulb or the House lead me back to the carousel, or the initial message? Is that where either of those lead? I found myself randomly tapping one or the other, and then stumbling around to find either the main story of the day, or the carousel of other potential stories and topics to browse through.</p><p>I‘m not a huge fan of the arrangement of this cluster of navigation elements. It seems like the Share button would not be as frequently used as the More button, and so giving it the high-value of the right-most position leads to a slightly less ideal user experience. Similarly, having the Question mark emoji (which leads to a Help card) in all summary messages is a bit much. While it’s nice to know that I can get help at any point, it seems like that would be better placed once I’m back at Home. Otherwise, the emoji nav items feel too tightly clustered.</p><h4>Keywords… not so key</h4><p>I had a confusing experience with keywords. Discovering that they ‘worked’ took several interactions with Purple. There was no on-boarding to suggest that this was an option to dive deeper into stories — or if there was, it wasn’t memorable enough that it held some understanding in my mind. I returned to Purple a couple times and was unaware that the ALL-CAPS keyword would trigger anything.</p><p>However, some of the messages from Purple called this out explicitly, and after a few of those interactions my experiences in using the product in that way informed my on-going use:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/537/1*PKcCizjMzK0LsodRArPLig.png" /></figure><h3>Sharing Purple</h3><p>Only recently has Facebook made sharing bots on Messenger a feature. I find Purple’s approach to be really cute, and very much in-line with the tone and personality of the bot:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/535/1*dENSTbHc4tgfvuogJJ-q0g.png" /></figure><p>There are also standard Share prompts at the end of some of the summaries (I couldn’t find a defining trait amongst those suggested to be shared), with a cute Flame emoji:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/438/1*VK1MyVOcEXjLr5GMO0VoYg.png" /></figure><p>Tapping on the Share element leads to a summary of the card, along with the option to share on Facebook, or a ‘Nevermind’ option:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/437/1*BiRbnNObJFFPE9zOF23FqA.png" /></figure><h3>The future of Purple</h3><p>I find Purple really compelling. The ability for the product to reach my core need — understanding what’s going on in politics today, easily and with just enough depth — is appreciated. The content is definitely the ‘shining star’ of my experience with Purple, and I enjoy reading through the various summaries that I find interesting. As you may have gathered from some of my notes above… much of the UI isn’t overly clear, and leads to some less-than-optimal experiences. The fact that I was willing to keep learning — and slowly get more comfortable with the difficult UI — is testament to the power of what can happen when you have a great core value proposition and are truly meeting the needs of your users.</p><p>However, I do wonder about the future of the product, especially as it relates to scaling the model as it is today. There have been a few occasions that hinted at this problem. For example, one of the cards in the carousel of stories (that’s still visible, as of the day of publishing this post) is about the VP debates… which was THREE whole weeks ago! Having compelling context and interpretation of news is one great feature of Purple, but ensuring that the content is fresh and compelling is something I consider to be table stakes for a product in this space.</p><p>As mentioned above, Human-in-the-Loop AI is compelling, but this isn’t quite what Purple is doing. There doesn’t seem to be any AI in the product. For example, typing a keyword in a not-quite-complete kind of way yielded no response from the bot:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/437/1*y4FMRpYhBpSOR5TU2xh8cA.png" /></figure><p>I also learned that using the keywords tied to the emoji navigation elements aren’t recognized by Purple either:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/438/1*4lZW_fWrerKgJrRhIhXZsw.png" /></figure><p>Similarly, queries that could be screened and filtered to respond to at scale don’t seem to exist, and Purple simply passes any and all queries onto either Rebecca or David:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/436/1*D3PGwwT6l-riJd_zcawkZw.png" /></figure><p>I appreciate the personal attention, but worry about what may happen to my experience once there are many more people constantly pinging the team or others who eventually take over that role. There are already signs of this type of delay impacting the quality of the product:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/437/1*fK6HIBj4OqLIbvCI8tTH3A.png" /></figure><p>If the Purple team can solve some of these problems, I’m excited about the future of how Purple delivers compelling news everyday, with the potential for me to get clarity or dig in whenever I’d like. If not, then I almost question if a conversational bot makes sense for this product, and if a website with static, older ‘news’, a focus on promoting events and collecting user feedback through polls, along with a ‘today’s hot off the press summary and insights!’-style newsletter may in fact be the more appropriate format ¯\_(ツ)_/¯</p><h3>Subscribe to the <a href="https://medium.com/ddouble">ddouble publication</a> to get every Tuesday Bot Teardown!</h3><p><em>ps: In case you didn’t notice… we’ve decided to move this series to a biweekly schedule. Me and </em><a href="http://medium.com/@micaelaneus"><em>Mica</em></a><em> are busy :) See you in 2w!</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=db25b53866df" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ddouble/is-the-election-confusing-you-this-little-bot-will-clear-things-up-db25b53866df">Is the election confusing you? This little bot will clear things up</a> was originally published in <a href="https://medium.com/ddouble">ddouble</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to make a context-aware bot: Foursquare’s big bet on Marsbot]]></title>
            <link>https://medium.com/ddouble/how-to-make-a-context-aware-bot-foursquares-big-bet-on-marsbot-c00d86d18eff?source=rss----f293534d269e---4</link>
            <guid isPermaLink="false">https://medium.com/p/c00d86d18eff</guid>
            <category><![CDATA[mobile-app-development]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[bot-teardown]]></category>
            <category><![CDATA[mobile]]></category>
            <category><![CDATA[messaging]]></category>
            <dc:creator><![CDATA[Sandi MacPherson]]></dc:creator>
            <pubDate>Tue, 11 Oct 2016 16:56:35 GMT</pubDate>
            <atom:updated>2016-10-11T17:20:13.233Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Vh0FhX9OLlxIZGQDvj_j3A.jpeg" /></figure><p>I’ve lived in Oakland for the past 2 years. Whenever I spend time in SF, a lot of the places are new to me — new neighborhoods, new bars and restaurants, new dog parks. Figuring out which ones to visit often comes down to opening up the Yelp app. However, that long-standing habit has recently been interrupted by Foursquare’s Marsbot. This unexpected behavior change these past few weeks prompted me to dig a little deeper, and do a teardown of Marsbot for the series this week.</p><h3>But first… A bit of background</h3><p>Foursquare burst onto the scene in 2009 at SXSW, when the annual event was more about music than tech panels and app launch parties. Two years later they had 10M users, with 35,000 new people signing up to the service every day. They raised a $20M Series B led by a16z in 2010, followed quickly by a $50M B (also led by a16z) one year later, valuing the company at $600M.</p><p>Unfortunately, the past few years have been a bit, shall we say… ‘tumultuous’ for Foursquare. Since launching over 6 years ago, the Foursquare app has slowly fallen down in the US ‘Food and Drink’, ‘Social Networking’, and ‘Overall’ charts:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IUi1SCpz5UM8A96Z0Ni-Hg.png" /></figure><p>As of this past Friday, the only category where Foursquare was in the top 1500 of any category was ‘US Food and Drink’, where it sat at #38.</p><h4>A ‘betting the farm’ level bet</h4><p>To combat this, and to clear up two distinct use cases for the product, CEO Dennis Crowley <a href="http://blog.foursquare.com/post/84422758243/a-look-into-the-future-of-foursquare-including-a">announced</a> a massive strategic shift for the company and their product. The original Foursquare mobile app would focus only on exploration and discovery, competing head-to-head with Yelp. Meanwhile, all of the social pieces of the product would shift over to a brand new product called Swarm, helping users find and coordinate with nearby friends, and do the ever-popular ‘check-in’ action the company popularized.</p><p>The split was a <a href="http://www.theverge.com/2014/5/1/5666062/foursquare-swarm-new-app">huge bet</a>. Looking at similar App Store data for Swarm almost 2.5y later, it’s unclear whether the strategy actually worked:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IHoXVa_IN4GSlXxNAqXp9w.png" /></figure><p>Swarm quickly fell off the US Overall rankings, and hasn’t cracked the top 100 in the Social Networking product for over a year (except for an 11d period this August, where it climbed to #89).</p><p><em>*quick aside… no, I’m not going to get into Foursquare’s other growth and monetization strategies via their data (i.e. Microsoft’s Cortana, Pinterest), just apps today, folks!</em></p><h3>Marsbot</h3><p>At the end of May, a small team at Foursquare <a href="http://blog.foursquare.com/post/144872708248/introducing-marsbot">released</a> a brand new product called ‘Marsbot’. The limited release targeted iPhone users in NYC and SF, and new signups were put on a temporary waitlist.</p><p>I got off the waitlist pretty quickly, and started using the product right away. However, not long after I went home to eastern Canada. Being out of the US caused Marsbot to stop contacting me (which makes sense — int’l SMS is $$$). But once I opened the app again randomly last week, Marsbot got back to work finding me great places to check out. Huzzah!</p><h3>So… what’s the vision?</h3><p>Marsbot is focused on pushing you suggestions, rather than being a product that a user opens with a pre-identified intent and then completes a query. Marsbot wants to surface great places to you before you even think you may want to check something out. Almost like that great friend who randomly messages you on a Thursday night, asking you to check out Cool New Place XYZ that’s totally on-point with your tastes:</p><blockquote>“We think of Marsbot as that best friend who knows what you love, where you hang out — and also happens to know all the best places in your city. Like my bestie, Marsbot learns that I’m into pistachio gelato and I spend most of my time in The Mission. Like my bestie, Marsbot texts me when it discovers a new chai spot opening around the corner from my office…” “…we are more focused about learning that you like pizza and proactively telling you about the new pizza place that opened up in your neighborhood.” -<a href="http://blog.foursquare.com/post/144872708248/introducing-marsbot">Marissa Chacko</a></blockquote><p>Instead of a ‘chatbot’, the Marsbot team thinks about the product as a ‘context bot’.</p><h4>But why download an app?</h4><p><a href="https://twitter.com/marissachacko">Marissa Chacko</a>, the Product lead on Marsbot, also <a href="https://www.producthunt.com/tech/marsbot">explained</a> a bit more about the experience her and the team were going for on Product Hunt, and why that necessitated an app:</p><blockquote>“…in order to have rich contextual data to personalize (location / identity / memory), we need an app (because Apple, Google &amp; FB won’t provide any other means) … developing bots on messaging clients will be cross-platform but at the moment still seems like very limited and would degrade the experience we wanted to provide (something beyond a question and answer experience). Ultimately we have to push the platforms on this one…”</blockquote><p>Marsbot is a prime example of the cluster of apps that are groups under the trend of ‘invisible apps’. As a user, you don’t actually open Marsbot for any particular use, except for if/when you want to adjust your settings. Users experience the product solely via SMS. This foundation makes the product seem a bit different from other bots — it’s rare for a bot to have a required app to download, but then interacts via the phone’s default SMS feature.</p><h4>Getting started with Marsbot</h4><p>The onboarding is super straight-forward and quick. The passive nature of the product is noted from the start, and setup is done with a few taps. One simple permission is all that’s needed:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*Ao8QrLHR_cVqK5kqZQn8oQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*AjnJtoksIulcWpzSfo4YTA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*Af8YIdoyOI_HPqX8b2imig.png" /></figure><p>The quick description of the product is simple and clear — Marsbot will find great things, without any work from you, and text them to you. Awesome!</p><h4>Getting to know Marsbot</h4><p>The onboarding for Marsbot is fully via SMS. After being on the waitlist for 3d, Marsbot texted me to get everything set up:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*fpwIx59q13vEbJhIT9D_1g.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/537/1*qFSioLWQ7zdqNI6X11z_dA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*JPlOo120AUl-lPxwHJu-Mg.png" /></figure><p>After going through this quick set-up, Marsbot adds this information (accessible via the app) and it begins tracking your location, understanding where you spend time, and the types of places you visit.</p><h4>Other product messages</h4><p>Marsbot followed-up with a bit of a tutorial 3d later, to explain how to use the product, simple commands that are available, etc., if/when the user chooses to complete some sort of ‘pull’ action in addition to the pushes they’re receiving from Marsbot:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/434/1*AsEod6IZVf9IeqLg8WMXiw.png" /></figure><p>One small critique — it would make Marsbot feel a bit more personal if the suggested search queries were related to me in some way. At this point, I’ve been using the product for almost a week. Marsbot should know at least a couple places I’ve been to and could potentially offer something more unique than ‘find cocktails nearby’, and could definitely suggest ‘find sushi in neighborhood-you’re-in-now/have-been-in-recently’ instead of a neighborhood I haven’t been in in over a year, on the other side of the country.</p><p>To confirm accuracy of the places that you visit over time, Marsbot will send a message with the details of a place that you visit often, asking if it’s in-line with your tastes and if you’d like to be notified of similar places as you go about your day-to-day:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/534/1*8bJwW74yox-y0H4VH7PpFw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*wUy5srYo1LPqXWeUWL-85w.png" /></figure><p>Marsbot also has a cute reminder of SMS fees, that I received after my on-going experience with the product had been running for a month:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/537/1*QO5xlD02hg0SCbK_hUanNQ.png" /></figure><h3>Marsbot in your life</h3><p>Being a Marsbot users means that you’ll randomly receive texts, most often suggestions on places to check out. Looking back at my text history, this is pretty typical of the types of messages you may receive as a Marsbot user:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/534/1*L2TQC75LIAcQx0yIp5VwnQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/535/1*RJxNy7IbcxFL9ypeoHZQJQ.png" /></figure><p>Out for dinner or drinks, and Marsbot acts as a bit of a concierge. Recommendations for what you might want to do next are surfaced, without any input or intent from the user. I’ve found the experience to be pretty seamless. It’s true that there have been a couple of occasions when I felt like Marsbot was interrupting me… but they haven’t become obtrusive enough (yet?) that I’ve turned them off.</p><p>Looking back at my text history, I’m unsure exactly how many of the suggestions I followed up on. I’m unsure if I keep Marsbot enabled, yet never actually ‘take Marsbot up’ on any of the suggestions if I’d actually stop the bot. Maybe?</p><p>There are no follow-up messages from Marsbot either, to ask if I ended up at any of the places that were suggested. But because they have all of my location data, it doesn’t quite seem necessary. They can make some assumptions around where I go next, time spent, and better feed that data into their suggestions algorithm to give me stronger recommendations in the future.</p><h4>Meaningful push messages</h4><p>One of the nuances about Marsbot that I appreciate is the choice of language in the messages. They’re often different, and the team has clearly done a great job deciding how to make the decision around which key piece of info will be shared with the user in the small, short space that’s available in one text. For example:</p><ul><li>“One of the top 5 places for live music in Oakland”</li><li>“I bet it has the cookies you’ve been waiting for!” (I told Marsbot I like cookies, lol)</li><li>“I hope you left room for dessert!”</li><li>“Looks like it might be your first time at Awaken Cafe”</li></ul><p>These are important cues, as setting the experience that the user will have and driving them to actually visit a physical place is a big ask. And if the person trusts in Marsbot and actually goes, I’m assuming they’d want the person to actually enjoy the venue and to have a good experience — as in the users mind, that recommendation is the only measure of quality for the actual Marsbot problem. So framing the place properly is extremely important.</p><p>Similarly, when you arrive at a location and are looking for suggestions, Marsbot has a chance to tie itself to a positive experience you may have there. It makes a ton of sense for them to surface what most people really enjoy, for example:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*20Tyxc7Z4YzJoI14A2DhEA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*iZQCuofsxJxQQ_EhHJDSKg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*M1kK_P3Rk7psjPW0s0tw7Q.png" /></figure><h4>Timing and context</h4><p>Another small detail that I really appreciated was Marsbot’s awareness of time and behavioral trends. It’s 10pm on a Friday? I just got to a bar? Of course it makes sense to follow-up with another bar that I might want to check out afterwards:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/537/1*XqD1g61uhkdbYy8z38h6sA.png" /></figure><h3>Chatting with Marsbot</h3><p>While the main use case of Marsbot means that you rarely actually interact with the bot, if/when it brings up something interesting, you have the option to dig in. Similarly, if you want to use the product as a Foursquare/Yelp replacement — and take advantage of the fact that Marsbot already knows about your preferred tastes — getting a recommendation is very straightforward:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/534/1*F1t-QnUmiZsETR3oOxwXNQ.png" /></figure><p>In this same interaction, Marsbot now knows that I didn’t take up the recommendation that had previously been served to me, and chose a different cocktail bar instead. In the tactic outlined previously, Marsbot doesn’t necessarily need to sit back while I ignore the suggestion. There’s an opportunity to put itself into my situation, and potentially tie the experience that I have at the location with Marsbot.</p><p>Even if there isn’t a lot of information about a certain location or type of place when I search for one (for example, I was looking for a Park in West Oakland one day), Marsbot doesn’t completely give up. There’s a suggestion, but it’s tempered with an apologetic, but still cute and in-character, response:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*cU88Xwu5UsUOuo8QMjnVdA.png" /></figure><h3>What’s missing</h3><h4>Travel</h4><p>One of the biggest gaps that I noticed was related to a trip I took to NYC, just a couple of weeks after installing Marsbot. I found it strange that Marsbot treated my time spent in the city as it had my time in SF, when I was in a very different place — not my home, I hadn’t been there before since starting to use Marsbot, etc. In my mind, I would have really appreciated a bit more of a ‘visitor’ type of mode, where things that may be more appropriate to spending that type of time in a city might be appreciated. At the same time, I do appreciate that Marsbot treated me basically like a local, and knowing my tastes, tried to recommend a place that I would probably actually enjoy.</p><p>This was called out more specifically while I was out of town another time. At this point, I had been using Marsbot a lot more, so perhaps it had collected more information and was able to offer a better suggestion… though I don’t think I’ve had Mexican food for over a month, so I’m unsure how it generated this suggestion:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/537/1*nXu_GnsFgngWlZ5_ByLCUg.png" /></figure><p>At 1:14pm, the lunch recommendation was appropriate and felt like a very supportive type of message to receive as I was in this new town for the first time.</p><h4>Non-food/drink suggestions</h4><p>If you choose to query Marsbot for a doctor, dentist, gym, etc., you’ll get a reply. However, it felt a bit strange using Marsbot for that purpose. Marsbot knows what kind of food I like, whether I like new and trendy places or not, the demographics of the people who like places similar to those I like, etc. None of that data is relevant (yes, that’s a sweeping assumption) when I’m trying to find a dentist, or a doctor, or a gym. Marsbot did send a response when I asked:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/537/1*w0QOgvx2lgjX42tlEap2Ew.png" /></figure><p>However, I almost wish there was a stronger product opinion that Marsbot isn’t built for this use case, reinforced through responses like this. For example, maybe Marsbot just straight-up says no, and doesn’t bother sending me random listings that are arbitrarily near me, or are related to me on some other tangential data. Getting a clear message from Marsbot about how it’s not built for that, and can only do certain things well, might be a stronger position.</p><h3>Technical: One marvel, and a few bumps</h3><h4>The marvel</h4><p>Marsbot does a great job at giving a follow-up recommendation if the first one isn’t quite a hit. For example, if I feel that a place is too far, a simple text of ‘closer’ prompts another recommendation to be sent that’s closer to my current location:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/535/1*4kAyf0bthOmDubjOfQ4Z-Q.png" /></figure><p>The ‘You’re welcome’ message is also a cute addition :)</p><h4>Technical bumps</h4><p>I ran into a few issues with Marsbot. Some of them really were small bumps, while others were actually a bit problematic:</p><ul><li>Typing ‘help’ doesn’t do anything in-app, and pushing me out to a different mode of interaction. While the best practices here for bots are still being developed, I felt a bit ignored by Marsbot in this case.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/534/1*7f8BYmgd2pKAVCu-edkaKA.png" /></figure><ul><li>I was picking up a friend in Eureka Valley, and Marsbot recognized that I was there — but my time in that neighborhood was simply transient. Pinging me at this point was one of the only pings that comes to mind where I actually found Marsbot to not be on-point, and to be a bit annoying.</li><li>When I asked for a sushi place, Marsbot recommended a ramen place that doesn’t serve sushi 🙃</li><li>Each SMS is prefaces with ‘ [marsbot] ’, which feels a bit janky</li></ul><h3>Final thoughts</h3><p>I’m unsure how strongly the Foursquare team is pushing Marsbot at the moment, and if it’s fully ready for the limelight. Stats from the App Store show that the app isn’t getting a lot of traction at the moment, and downloads have fallen off since the release in the spring:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gLTAXWNZNe9aMAPfCJ7gzQ.png" /></figure><p>It seems as if it’s more in an experimental phase at the moment, and being limited to SF and NYC would also have a huge impact on the potential for the product to achieve broader adoption.</p><p>I am curious to see if or when the team decides to roll out support for other types of recommendations. Personally, I believe that the frequency of all food and drink activities are probably the only ones that make sense for the product. Sending notifications for any other type of service would be very infrequent, and the data that Marsbot is collecting is probably not related to my preferences and decisions in other categories. For example, I don’t get my hair cut multiple times per day, so having Marsbot send me recommendations of great hair salons doesn’t make sense in the context of a context-aware SMS bot, and where I ate dinner and got drinks this past week is probably a difficult dataset to work with to suggest a hair salon. As mentioned previously, I hope that the product becomes more tightly focused and potentially even removes some of the responses that it’s currently serving for those types of queries.</p><p>One big win for Marsbot is that it’s definitely cut into my Yelp usage, which I’m assuming is one of the main goals of the product. There was one clear case where I was thinking about somewhere to head to next, and I followed the suggestion that Marsbot offered. It would be interesting if I had better tracked my activities and which apps I was using at which times, to understand when I defaulted to Yelp — both as a way to measure the success of Marsbot, and also as a way to better understand when they could better ping me with suggestions. For example, I’m assuming some people are more planners, and prefer to get pinged about suggested places for drinks on Friday night in the mid-afternoon — but because they go out for drinks EVERY Friday night, in the same collection of a few neighborhoods, the slight time-shift would be appreciated.</p><p>The power of Marsbot to collect information as I go about my life is also one point that I find extremely powerful and interesting, and one that I (and Foursquare!) hope will only make my experiences with Marsbot better and more fulfilling over time.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c00d86d18eff" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ddouble/how-to-make-a-context-aware-bot-foursquares-big-bet-on-marsbot-c00d86d18eff">How to make a context-aware bot: Foursquare’s big bet on Marsbot</a> was originally published in <a href="https://medium.com/ddouble">ddouble</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How I came to love an annoying millennial: Bots are people too]]></title>
            <link>https://medium.com/ddouble/how-i-came-to-love-an-annoying-millennial-bots-are-people-too-3d25fc2a14d8?source=rss----f293534d269e---4</link>
            <guid isPermaLink="false">https://medium.com/p/3d25fc2a14d8</guid>
            <category><![CDATA[gaming]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[bot-teardown]]></category>
            <category><![CDATA[messaging]]></category>
            <category><![CDATA[product-design]]></category>
            <dc:creator><![CDATA[Sandi MacPherson]]></dc:creator>
            <pubDate>Tue, 04 Oct 2016 16:45:00 GMT</pubDate>
            <atom:updated>2016-10-04T16:44:59.325Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/930/1*NcTb6y4HwLEMhvcyfRnkjw.jpeg" /></figure><p>Compared to the never-ending abyss of “do sh*t for me” bots, Jessie stands alone. A fully developed character that you can interact with over Messenger, Kik, or Skype, Jessie will take you on a journey through the day of a woman stumbling through life as a 20-something in LA. The experience is much more game-like than bot-esque, and hints at one of the many potential futures where scripted conversational-based agents can shine. <a href="https://medium.com/@micaelaneus">Mica</a> and I interacted with Jessie this past week, and gleaned a ton of insight into how this type of bot is designed, what makes Jessie ‘tick’, and what the future of these types of bots means for people building any kind of bot.</p><h3>The industry’s best invents the future</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qPtMYyoHITBEWw3bpzQ5aA.png" /></figure><p>Jessie was built by the team at Pullstring, previously known as ToyTalk. Founded by 3 (!) ex-Pixar folks (including CEO Oren Jacob, previous CTO of Pixar), Pullstring began working in the scripted character/bot space a few years ago. They’re responsible for the talking <a href="http://hellobarbiefaq.mattel.com/">Hello Barbie</a> from Mattel, the <a href="http://mashable.com/2016/05/01/call-of-duty-codmessenger-arg/">Call of Duty</a> Facebook Messenger bot, and the Thomas the Tank Engine ‘<a href="http://www.thomasandfriends.com/en-us/products/apps">Talk to You</a>’ app. In addition, Rod Humble (ex-CEO of Second Life’s Linden Lab, ex-SVP at EA on The Sims, and now GM at SGN Games) was most recently Chief Creative Officer at Pullstring, and was brought on to build Jessie. I don’t think anyone could assemble a stronger team to build the future of character-based bots.</p><p>Chatting with Jessie pulls you into her world. She responds almost as if she’s a close friend, asking you for your opinion on how she should spend her day, whether she should abandon a friend who she’s promised to help move, and updates you when she has to step away to order a latte. Interacting with Jessie doesn’t necessarily ‘do’ anything for you. The product is pure entertainment, exploring the potential of messaging and chat as the delivery mechanism.</p><p>Released in early April 2016, the main character Jessie was described by the Pullstring team as:</p><blockquote>“…a ton of fun, but a bit of a hot mess. She is a quirky, energetic 20–something experiencing a quarter-life crisis and looking for a job, a boyfriend, and a new apartment — all in the same day.”</blockquote><p>As you interact with Jessie over her day, you have the potential to navigate through one of nine possible outcomes. The total ‘product’ consists of over 3000 lines of scripted dialogue across those nine stories, one of which takes anywhere from 25–30 hours to fully navigate, from your first ‘hello’ to some sort of conclusion.</p><h3>Meeting Jessie</h3><p>After adding Jessie to Messenger, you’re greeted with a simple welcome card, prompting you to ‘Start’ the conversation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/824/1*8guKUQr23GLeHCQZqx7v7Q.png" /></figure><p>All of the standard indicators exist — the small profile photo as a read receipt indicator, typing indicator bubbles, and a link to her Facebook Page.</p><p>As with all products, one of the most important actions during the onboarding flow is to set the context of the interaction, and to define what and how you’ll be using the product. With a character-based messaging bot like Jessie, this means that the actual words used are extremely important, along with how the bot exposes their world and circumstances. For Jessie, this means immediately describing what she’s up to, and the type of journey you’ll be going on with her:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/513/1*q3Sicq2VUwe-H3RHhEQk3A.png" /></figure><p>In addition, as you can see, the tone of her dialogue is extremely powerful. It’s obvious that Jessie is opinionated, but fun. The word choice is very descriptive of the character — she’s a funny, young, playful woman that doesn’t feel the need to be concise in how she interacts with people. Establishing this as Jessie’s personality is extremely powerful later in the experience, and was a smart decision as a way to design around some of the current gaps/drawbacks/weaknesses of creating character-based bots (which we’ll explore later in this post).</p><h3>One (of nine) stories</h3><p>While I wish we had unlimited time to review bots… we don’t. I only had the chance to go from start to end of one of the nine stories. I mostly tried to ‘play along’ and interact with Jessie as I expected an early 20-something woman would (my assumption as to their target user), but did ‘poke the bear’ a few times to try to better understand how Jessie was built, as well as to explore what might happen if and/or when a user deviated too far outside the established domain.</p><p>Jessie feels very much like a ‘choose your own adventure’ experience. Once you make a decision as to how Jessie should act, you’re set down a certain path that you’re then pushed along. You hit various decision nodes along the way (i.e. stay in for the night, or go to the CRAY CRAY yacht party!!), and continue to be shuttled along your chosen path.</p><p>Here’s a summary of what happened in my conversation with Jessie:</p><ul><li>After we say hi, I learn Jessie is at a cafe, looking for a new apartment.</li><li>Finds a decent apartment, and I help her fill out the application.</li><li>She goes to the open house, and tells me about her potential roomies.</li><li>Doesn’t get the apartment :(</li><li>Jessie asks for my advice for an interview she’s on her way to.</li><li>I help her navigate an awkward situation with a guy in the elevator on her way to the interview… who turns out to work at the company!</li><li>She messages me during the interview and I help her answer questions on the interview test (!).</li><li>Sleepy, she goes home after the interview and takes a nap.</li><li>When she wakes up, the guy she met in the elevator invites her to a party on a boat. I tell her she should go, vs. stay home and eat pizza.</li><li>I help her pick out her outfit for the party.</li><li>She meets a guy at the party who gives her a $10k chip for a casino nearby.</li><li>She heads to the casino, and ends up accidentally buying a $20k bottle of champagne.</li><li>She heads to the roulette table to try to make back the $10k she’s down…</li><li>The story ends.</li></ul><h3>Interacting with Jessie</h3><h4>Jessie acts like Jessie</h4><p>Interacting with Jessie felt a lot like interacting with what I imagine interacting with a 23 y/o woman, based in LA, would feel like. When she told me that a guy in the elevator with her had ‘bird 💩 or gum or SOMETHING’ on his pants, and I asked her for a photo — she sent one right away:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/516/1*OGguNR2I5blZJi0NqXT3vA.png" /></figure><p>Jessie’s other actions and behaviors mirror those of what I’d expect. For example, she sends me a message during her job interview (!), asking me to help her:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/515/1*uxrO-4F8TOZuAQp3nLrwtQ.png" /></figure><p>Another layer of ‘real-ness’ is added to the character through the content that’s shared in this exchange. She jokes about sex, and having that as a topic in the conversation in a non-intimate way helps to also tether the conversation to a topic that’s very common for her demographic. It has the added benefit of further creating a sense of connection with Jessie, as she’s comfortable bringing up a topic that’s often only reserved for close friends.</p><h4>Real people aren’t instantaneous</h4><p>Timing is one variable that bots can adjust, and it’s obvious that it’s something Jessie’s creators spent a lot of time thinking through. The response times from Jessie are varied, based on the situation. If she starts down a conversation thread with you she doesn’t suddenly become unresponsive, but she does sometimes hop out of a current conversation, and goes into an unreachable state.</p><p>For example, as Jessie is wrapping up her job interview, she lets me know that the interviewer is returning and she has to put her phone away:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/1*l9xDG24zO3b5iXsqEuZ9lA.png" /></figure><p>The gap between my ‘good luck!’ message and her ‘:D :D :D’ message was a full 4 minutes — what you’d expect when someone wraps up a job interview, says goodbye to the interviewer, leaves the room, and has a chance to take their phone out again.</p><p>Jessie takes many similar breaks between messages, all of which vary in length. I experienced breaks of 30s, 1m, 3m, 4m, 7m, 10m, 15m, and 60min in my conversation with Jessie. All of them felt appropriate considering the situation. For example, there’s a 1 minute break between this ‘You around in a few mins? I may need you!’ message, until she gives up on waiting for me and sends an ‘Oop! It’s starting! Text ya soon.’ message:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/515/1*lyYJoaVDfNLOMpcI_gUecg.png" /></figure><p>This wasn’t always perfect, however. At one point near the end of our conversation, Jessie was at a party. I was busy and didn’t message her for over 24h… yet when I did, she was still at the party, hanging out with the same people as when I’d left.</p><h4>Answering with non-answers</h4><p>The Pullstring team has done a great job at getting Jessie to always reply in a way that’s realistic, even if it’s not quite appropriate. At one point I ask her to send me a photo of the ‘hottie’ she spotted at the cafe:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/515/1*03JUiz4WbDkbZaON8-WdBw.png" /></figure><p>Instead of replying with a default ‘I’m sorry, I don’t understand’, or a ‘No photo exists’ type error message, Jessie’s response doesn’t quite respond to my request, but it’s a reasonable response nonetheless. It’s a response that I’d expect a friend to send, further increasing my sense of Jessie as a real person.</p><h4>Language</h4><p>As previously mentioned (and has probably become obvious via the screenshots), the language that Jessie uses throughout our conversation is extremely powerful in establishing the richness of her character and placing her in my mind as a real person, with a real age, a real personality, a real existence. Some of my favorite words that Jessie uses are:</p><ul><li>gunna</li><li>killa</li><li>lol</li><li>tho</li><li>bajillion</li><li>WDYT</li><li>prolly</li><li>ermagawd</li></ul><p>She also uses plenty of emoji — both as responses and mid-sentence — to either replace words, or express emotions. The one similar piece of ‘language’ that I would have expected but isn’t used is gifs.</p><h4>People don’t respond directly sometimes</h4><p>One interaction that I really appreciated, and that leant itself especially to having Jessie message as an actual person would, was this one:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/515/1*gvYBfD-ageKRKX711kDEkQ.png" /></figure><p>Personally, I do a very similar thing when messaging. If someone asks me a question, and I need to somehow confirm that I’m still there but unable to answer immediately, I’ll indicate that. Jessie’s ability to recognize that my ‘ummmm’ message wasn’t a response, but simply a placeholder until I could come up with an actual response, was a very strong design decision.</p><h4>Pulling you along</h4><p>The dialogue does a fantastic job at keeping you on-track, and not really letting you drift outside of Jessie’s overall domain or the story thread you’ve started down. You’ll be re-prompted to answer a question if you give a non-answer, a duplicate, or one that doesn’t lead to the story line that Jessie is targeting. When I answered with basically the same thing, Jessie noticed and called me out directly:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/517/1*9oYFjxMIIEzqbRxere1UZw.png" /></figure><h3>The power of Jessie’s millennial LA persona</h3><p>I have to be honest — I don’t actually love Jessie, I sorta hate her. She’s obnoxious, extremely self-centered, and can’t maintain focus on anything (other than a hottie she spots in a cafe) for longer than 5ms. When she would ‘come back’ after taking off to do something during our chats, my most common reaction was a deep sigh of relief.</p><p>That said, the persona of Jessie is genius. She’s constantly distracted, only wants to talk about herself, doesn’t quite give what you say any credence — which makes any of the bugs or rough edges you may encounter blend seamlessly into the character.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/517/1*S5X2x5SvzjVwCS9Zzw0kRg.png" /></figure><p>These ‘non-response’ responses are brilliant. Responding with a high degree of accuracy to any potential message is a very difficult tasks at the moment, considering the current state of the technology that exists in <a href="https://medium.com/ddouble/how-to-build-your-best-bot-the-bot-stack-compendium-90a90660167a">the bot stack</a>. By using this detached, self-centered, scatter-brained persona, it’s basically expected that Jessie will drive all conversations back to her neediness and inability to care or engage you with anything (i.e. literally anything, an infinite domain) going on in your life.</p><p>The fact that Jessie exists almost exclusively inside of time-boxed situations also means that there’s always a reasonable excuse for her to jet off — the party is starting, the apartment showing is starting soon, her job interview is over, etc. This allows her to jump out of the beginning of any conversation that has the potential to become too complex for a (smart, AI-enabled) scripted bot to manage.</p><h3>Design Decisions for the Messenger interface</h3><p>A quick note here… While Jessie is available on Kik, Skype, and Messenger, I only interacted with the Messenger version of the product, so can only comment on the design elements used on that platform.</p><h4>Read Receipts</h4><p>The most obvious and frequent design element that I noticed was the use of Messenger’s read receipts. When the bot has ‘read’ a message you’ve sent, Jessie’s mini profile pic moves down to that line, to show it’s been seen. However, for many of my messages — most often those that were sent after Jessie said goodbye — the ‘read’ status was postponed until after Jessie returned. It made it clear, in a subtle way, that Jessie was unavailable, and that she had dashed off and hadn’t yet looked at any messages that had since come in:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/1*T1ofrFI54awfirtCasKtsQ.png" /></figure><h4>Messenger’s Cards</h4><p>The team at Pullstring uses Messenger’s cards in various points throughout chatting with Jessie.</p><p>The very first card in my conversation with Jessie was sent a bit abruptly. Jessie had taken a break, and a card with a Share CTA was inserted into our conversation:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/509/1*L0UWAe2k2dO9iXfsorEBtg.png" /></figure><p>Other than Jessie deciding to step away form the conversation for a minute at that point, there was no other obvious indication as to why this point in our conversation was chosen as the ideal time to insert a share CTA.</p><p>Later in our conversation (and the only time I encountered it), the carousel element is used so I can help Jessie choose which outfit she should wear to the party:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/514/1*Jef2FMry7vW0eKIZAybN_w.png" /></figure><p>The carousel displays 3 possible outfits for you to choose from, as Jessie gets ready to go to a party on a yacht. The capacity for the carousel to offer options on visual information is clearly a great match for this situation, and helps keep the conversation interesting by switching to a completely different, non-text based interaction that may also require a physical change — either a tap or mouse click on the screen.</p><p>Once you’ve come to the end of your story, a final card is displayed:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/519/1*mIQPpxU07XeA3qdMZT__uQ.png" /></figure><p>Not choosing either Share or Play Again displays a slightly different card, with options to learn more about the story or Pullstring:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/516/1*08AKo2NfXas-gUZWXbaEwA.png" /></figure><h4>Visuals</h4><p>One of my favorite parts of interacting with Jessie was the illustrations that were shared throughout our conversation. These ‘pictures’ of what was going on around Jessie really helped me to imagine her interacting with the real world. Excluding the outfits I chose from the carousel, I received a total of 5 illustrations during our chat:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/287/1*_qIjGzew103Z4oRngbiY5g.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/294/1*vciA7_eGHoVZ4q2mkyaR3w.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/434/1*V0DM68QCJY70NqurXXXHhw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/287/1*PCGgnaJeRJDkFzb0-4V4aQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/432/1*bqnsCcisE36GpJ3R1fWd4g.png" /></figure><h3>The state of character bots</h3><p>Interacting with Jessie was very different than any of <a href="https://medium.com/ddouble/tuesday-bot-teardown-growbot-ff997e785582">the</a> <a href="https://medium.com/ddouble/how-the-white-house-made-a-bot-out-of-an-empty-box-745441743122">other</a> <a href="https://medium.com/ddouble/how-to-get-teams-addicted-to-your-slack-bot-a-teardown-of-howdy-f22176149be0">bots</a> that we’ve reviewed. Jessie’s purpose is to entertain, by establishing a real, personal connection. The folks at Pullstring believe that these types of bots are the future of entertainment.</p><p>With Pullstring’s <a href="https://techcrunch.com/2016/09/27/pullstring-bot-authoring-tools/">recent launch</a> of their bot creation platform, the team clearly believes in this type of interaction as one possible (large) future that entertainment will take. I’m sure the team’s background in gaming will be leveraged as they continue to build out compelling storylines and worlds for their future characters.</p><p>As the tools for building these types of bots develop, it will also be interesting to see if or how the platforms where these games/experiences/chats live change. Do read receipts make sense for a bot? Is it okay to show that Jessie is active on Messenger… yet she’s not replying to my messages, or she’s told me she’s taking a nap? Does the metric of ‘Typically responds in minutes’, a metric associated with Facebook Pages (which is required to have a Messenger bot) make sense to display alongside a character?</p><p>While I didn’t personally enjoy the content that I interacted with via Jessie, I developed a deep respect for the carefully crafted situations. The normalcy of the interaction was very, very consistent — it was rare that I felt jarred from an underlying sense of interacting with something kinda-sorta like a person. And there were many moments where Jessie actually did feel real, or caused my mind to stir up memories of a friend who had once been in a very similar situation to one that she was navigating.</p><p>While the “do sh*t for me” genre of bots are distinct enough from actual people that how or when they deviate from ‘real’ people is obvious, the fuzzy gray areas that character bots like Jessie will expose will lead to some of the more compelling design decisions over the coming months as it relates to all bots.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3d25fc2a14d8" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ddouble/how-i-came-to-love-an-annoying-millennial-bots-are-people-too-3d25fc2a14d8">How I came to love an annoying millennial: Bots are people too</a> was originally published in <a href="https://medium.com/ddouble">ddouble</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to get teams addicted to your Slack bot: A teardown of Howdy]]></title>
            <link>https://medium.com/ddouble/how-to-get-teams-addicted-to-your-slack-bot-a-teardown-of-howdy-f22176149be0?source=rss----f293534d269e---4</link>
            <guid isPermaLink="false">https://medium.com/p/f22176149be0</guid>
            <category><![CDATA[slack]]></category>
            <category><![CDATA[management]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[bot-teardown]]></category>
            <dc:creator><![CDATA[Sandi MacPherson]]></dc:creator>
            <pubDate>Tue, 27 Sep 2016 16:33:44 GMT</pubDate>
            <atom:updated>2016-09-27T16:33:44.397Z</atom:updated>
            <content:encoded><![CDATA[<p>Howdy is a friendly Slack bot that runs your team’s meetings. Howdy was early to the Slack platform, and has many (well deserved!) fans amongst the startup crowd, especially as it relates to running the common ‘standup’ morning meeting. It’s the one bot that took basically zero effort to include in our daily routine, and has become one of our favorite bots.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/684/1*xSQ5zycnidF3Iv1PDwhhYA.png" /></figure><p>Howdy is built by the <a href="http://xoxco.com/">XOXCO</a> team in Austin, Texas. They’re the same team that’s responsible for the popular bot framework <a href="https://howdy.ai/botkit/">Botkit</a>. Out-of-the box Howdy will run a daily team checkin, collecting information from each teammate and collating it into a clean summary. But this is just the start, according to Howdy’s Founder and CEO Ben Brown:</p><blockquote>Our big goal is to make a bot that anyone can customize to be _their_ bot, so that all teams can enjoy the benefits of automation and digital assistance without having to, you know, build a bot. — Ben Brown</blockquote><p>It’s a bold vision — sharing the power of bots with anyone working on a team to help them work more effectively and efficiently. Considering the uptake and our experiences in using Howdy, they’re off to a great start. Let’s check out some of Howdy’s special sauce, and what makes the product both useful and sticky.</p><h3>Hello Howdy!</h3><p>The basic interactions with Howdy are simple — in the best way possible. Managing the bot(s) that your team is running is very straightforward, with a simple dashboard on the Howdy website to make any edits to your bots:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/829/1*foZDIyMXdgTHUaO8b2U2IA.png" /></figure><p>We began using just the defacto ‘checkin’ script for our team. To get started, you simply set the time, days of the week, where you want the final report of the meeting to be posted, how long Howdy should wait to collect responses, and the names of the participants you want included in the recurring meeting:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/829/1*yPKShfsGL1NKLcsH33Mjzw.png" /></figure><p>Me and <a href="https://medium.com/@micaelaneus">Mica</a> stopped there, and didn’t make any other tweaks to our script. We don’t need anything more customized than the standard 3 questions that most people at startups today answer as they run their morning standups.</p><h3>Talking to Howdy</h3><p>So what does it look like if you’re using Howdy for your standups? At the scheduled time, Howdy DMs each person on your team to remind them that the checkin is starting:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/550/1*gOC6Jxh4RcTI_pNAlr245A.png" /></figure><p>Once you click the ‘Begin’ button, Howdy launches into the script.</p><p>(<em>As a side note… Howdy used to ask you to type a simple ‘y’ or ‘yes’ to start running the meeting. The button UI is new as of a couple weeks ago</em>)</p><p>If you’re having a bit of a busy day, Howdy gives you a 5min warning before the meeting is scheduled to end so you can quickly add your notes before the meeting wraps, so they can be included in the daily summary:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/553/1*G39yLxj0lLbxGI1zZndIGQ.png" /></figure><p>This message is often appreciated… but for some reason it’s not sent consistently. Some days we receive this notification, but not all days — unsure if this is a feature or a bug!</p><p>Once you accept and start the meeting, each line of the scripts runs and waits for your response in your Howdy DM conversation:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/549/1*uptDou2BNfom013xrtE3HA.png" /></figure><p>If you’re too late and try to answer the prompts after the designated end time of the meeting, Howdy sends along a polite ‘Sorry, this meeting is already over.’ message to let you know that you’re unable to add your points any longer.</p><p>After the allotted duration of minutes/hours runs out, Howdy posts a summary of the notes, so everyone can roll their eyes over what the rest of the team is working on. The report clusters the questions, and shares the response from each person, posted directly to the designated channel. A couple of links are also included — you can review your own notes, and also edit the script or schedule:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/489/1*iPicDyn5wTAKdD6snVK02A.png" /></figure><p>All of these notes are collected over time, and are accessible via the Howdy website:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ciamy6yBwXYGuiWOLHh2nQ.png" /></figure><h3>Make it your own</h3><p>Creating a new script is super simple, props to the Howdy team for making what could be a complicated process into one that feels like something wholly non-technical, and all wrapped in a friendly UI too. Starting from the Scripts dashboard, you simply identify what exactly you want your bot to ‘listen for’, i.e. what the keyword will be to run your script:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/830/1*eH42KYzl5afzSYH4RHlOtg.png" /></figure><p>I really like this language, the use of the ‘I’ pronoun is a powerful and compelling choice. In naming the script in the way that personifies the bot that will run it, the very first step that the script-builder takes is one where the bot they’re building feels as much like a living, breathing digital coworker as it possibly could. 🙌</p><p>Once you’ve ‘named’ your bot by defining what keyword it’ll listed for, the Editor view allows you to build out the back-and-forth statements that your bot will run during each meeting.</p><p>You can create a back-and-forth dialogue of the questions if you’d like included via the simple, clickable dialogue view:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/723/1*dW5jNFFyaA2_Jy5NR5yMPA.png" /></figure><p>Simply type each message as you want them to appear:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*wC1xtOPIgWllw7DFoviBhA.png" /></figure><p>And the line will be added to the flow. Each new script entry has the option to be either closed or open ended, via a simple ‘Is this a question?’ checkbox — which is, again, a very clear way to confirm the type of action the bot should take in a non-technical way, completely devoid of the potential code that’s being triggered by the GUI that’ll drive a new row in the database (or not!):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/332/1*J-QZadi3xa2G7wADG3TkOg.png" /></figure><p>If the new line of the script is a question, the editor automatically displays a simple response bubble image, along with an explicit message stating that the bot will collect and store the information that’s entered when the script is run:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*b8wIvmaVYohXLWHmfea8mw.png" /></figure><p>You can create scripts that don’t have pre-defined participants, which could be useful for a certain format of meeting that has rotating attendees. In those cases, you can call the script via a simple ‘@howdy run script_name’ command in Slack in the designated channel, and Howdy will follow up with a quick question on who should be invited to participate:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/418/1*x3mkQYjO1xpAEPqXyPsGhg.png" /></figure><h3>Why we love Howdy</h3><p>The current thinking around bots is that those that do well are ones that have a very narrow use case, and are basically form replacements. Howdy definitely checks both of those boxes. Instead of a boring form to run your startup (which no one would actually do, anyway), participants interact in a conversational way around familiar questions. It fits really well with the existing behavior of users, and in doing those established behaviors via the bot, brings a ton of added value to the interaction — responses are automatically recorded, reminders are sent if people are late or forgetful, cute responses to answering questions that are meaningful to your workday.</p><p>It’s these subtle shifts in perception that we found caused us to most appreciate and value Howdy. Originally, we didn’t really have high expectations for the bot (#truth), and thought it would seem overkill, or we’d just never bother using it. But it has quickly become a routine just because of the fact that the over-arching practice is already in our routine, there was a negligible amount of user education to get us up to speed, and we’ve had literally zero friction or drawbacks since jumping aboard the Howdy train.</p><p>I’m unsure how other teams use Howdy (post-standup? pre-standup? how long before/after? etc.), but we use it 5min prior to our irl standup. It works because it has the same effect as Amazon’s <a href="https://conorneill.com/2012/11/30/amazon-staff-meetings-no-powerpoint/">6-pager approach to meetings</a> — the structured prep-work means that each individual comes to the standup having done enough work on their own to be optimized for interaction with the team. Howdy causes this shift in behavior, so while that change sits outside the chat bubbles of the bot itself, the positive outcomes and value derived from that shift have been 100% attributed to the bot.</p><p>While we haven’t created any personalized meeting scripts that we use, the potential to create other meeting formats is a powerful feature that we didn’t expect, and believe would be useful for other teams with a wider range of structured meetings across their calendars.</p><h3>Quirks we put up with</h3><p>We struggled a bit to come up with what we thought were ‘problems’ with Howdy. Partially because it’s such a simple product, but also because we haven’t run into many issues since starting to use it everyday for the past month.</p><p>One small issue is that sometimes it’s unclear how to ‘wake up’ non-scheduled scripts. Simply typing the name of the script doesn’t work — you need to both call Howdy, followed by the word ‘run’, and then the script name:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/590/1*srpYRafGXJCiLIMedz2kjg.png" /></figure><p>At the same time, we appreciate that Howdy’s Founder and CEO perspective on privacy:</p><blockquote>Howdy only pays attention to messages directed at the bot itself — no other messages are logged or stored, including any messages sent before the bot joins. — Ben Brown</blockquote><p>A recent change that we don’t like is the addition of buttons to the daily ‘are you ready?’ DM that Howdy sends to each participant. The action that you’re about to begin with Howdy is one where you’ll be typing on your keyboard, but tapping on a button requires the user to recognize the visual UI of the button, and then physically leave the keyboard, reaching out for their mouse in order to get started. Both of us much prefer the simple action of typing a short command to then launch the rest of the typing-based interaction.</p><p>One other small thing is that the ‘you’re too late, meeting is over’ message doesn’t fit with the otherwise friendly, engaging experience of using Howdy. Having a feature that would allow you to edit your submissions — even if they were noted as late submissions, with a timestamp of when they were added — would be great. For our purposes, when we miss a meeting, we manually log the 3 questions into our Standup channel. While they’re now in that location for future reference if needed, they’re not included in the Notes that Howdy otherwise dutifully logs everyday. And unfortunately for Howdy, when we look back on our notes, we’ll probably only blame the product for it’s lack of data (versus our tardiness!).</p><h4><a href="https://medium.com/ddouble">Follow the ddouble.ai publication</a> to make sure you don’t miss any future Tuesday Bot Teardowns!</h4><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f22176149be0" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ddouble/how-to-get-teams-addicted-to-your-slack-bot-a-teardown-of-howdy-f22176149be0">How to get teams addicted to your Slack bot: A teardown of Howdy</a> was originally published in <a href="https://medium.com/ddouble">ddouble</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How the White House made a bot out of an empty box]]></title>
            <link>https://medium.com/ddouble/how-the-white-house-made-a-bot-out-of-an-empty-box-745441743122?source=rss----f293534d269e---4</link>
            <guid isPermaLink="false">https://medium.com/p/745441743122</guid>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[government]]></category>
            <category><![CDATA[bot-teardown]]></category>
            <category><![CDATA[facebook]]></category>
            <category><![CDATA[messaging]]></category>
            <dc:creator><![CDATA[Micaela Neus]]></dc:creator>
            <pubDate>Tue, 20 Sep 2016 18:39:11 GMT</pubDate>
            <atom:updated>2016-09-20T18:39:10.892Z</atom:updated>
            <content:encoded><![CDATA[<p>The White House bot is boring, creepy, and reads like a rough draft. Still, it offers some insight into collecting personal info with a bot.</p><p><a href="http://medium.com/@sandimac">Sandi</a> and I mostly look at a lot of bots where the product is the bot. In this case, the “bot” is better described as an example of a conversational UI, a means to an end that have nothing to do with ongoing dialogue between two parties.</p><p>This bot is the functional equivalent of a cardboard box with a slot in the lid and <em>Suggestions</em> sharpied on the side. You’ve seen a version of this box, probably sitting on a table in a corner somewhere, and you know how effective it is. (Not very.)</p><p>Can we expect more from this bot than we can from an empty box?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/930/1*sPrekLRfLo463x6GitKAJw.png" /><figcaption>They announced the bot with a banner pic of the POTUS on a landline lol</figcaption></figure><h3>Why a White House bot?</h3><p>The core purpose of the bot isn’t to be a bot per se. It serves the more general goals of keeping the White House up-to-date with modern communication and letting citizens send feedback to their President.</p><p>Obama’s Chief Digital Officer Jason Goldman wants the White House to <a href="https://www.whitehouse.gov/blog/2016/08/10/now-you-can-message-president-obama-facebook">“meet people where they are.”</a></p><p>Now, is this an effective form of political engagement? Please discuss this question anywhere except in the comments below.</p><p>Their choice is a strong comment on the perceived value of conversational UI and by extension both bots and messaging apps as platforms. If bots are a fad, then so are Twitter and <a href="https://techcrunch.com/2016/06/22/michelle-obama-is-now-on-snapchat/">Snapchat</a>.</p><p>You will find the bot on Facebook Messenger. I see no word about future releases on other messaging platforms.</p><h3>Interaction and design</h3><p>The bot is meant to replace <a href="https://www.whitehouse.gov/contact">the existing web form</a> with something more social, less intimidating. A fair comparison should weigh its merits against those of the current solution.</p><p>So let’s look at the form:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/632/1*rFkuoRnMssS3pTNOba3FNw.png" /><figcaption>This forms lives on a website that you will never willingly visit.</figcaption></figure><p>It’s not a very good form, you might notice. It starts with the tedious identity fields (all but two marked Required) before asking what it was you wanted to say in the first place.</p><p>The bot gets an easy win by switching to the friendly interface of chat. But how different is the experience, really?</p><h4>Does having split personalities count as structure?</h4><p>The White House delivers both the best and the worst of bot design with the efficiency and bland cheer of a White House tour guide. Also, that guide is sizing you up as a prospective terrorist.</p><h4>First, meet a relaxed and energetic note-taker</h4><p>The bot leads with the value prop and good dose of energy:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/676/1*V73Y0zTzSxeAcIRdvLuFMA.png" /></figure><p>The language is… basic? The bot strikes an informal, upbeat tone which is a welcome change from the sober gray and baldly practical web form. Yet somehow it carefully steers away from expressing any individuality, speaking as some anonymous “we” of an organization.</p><p>Who am I talking to, a subcommittee? The interns?</p><p>Other bots that use buttons or suggested responses have felt a little awkward or prescriptive to me, artificially limiting my interaction with the bot. Here I found them appropriate, perhaps because the tone is so impersonal.</p><p>Messaging platforms will release more and more of these interactive elements. My experience suggests that we should use them judiciously due to their effect on a bot’s “feel” as much as their success in getting the desired user response.</p><h4>Editing your message to the White House</h4><p>After collecting your message for the President, the bot asks for confirmation before sending it. We all say things in anger that we later regret, things which impact the relationship forever. And we can’t spell good.</p><p>Except you can’t actually edit the message. You can append another sentence or two (Facebook sets a character limit on messages), or you can start over.</p><p>Starting over means you return to the original welcome message as if you have never previously interacted with the the bot. I returned to the chat a few days after my initial chat and found that you once again “start over” even if you’ve sent a message successfully in the past.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/668/1*1UqtcnT75NhANR-Lkt_1bg.png" /><figcaption>Editing means starting over from scratch.</figcaption></figure><p>No technical limitation requires starting over, really. It’s a design choice and not a strong one. The web form doesn’t remember your name either but it does allow for editing a message.</p><h4>Now time to meet the interrogator</h4><p>But okay, I’m satisfied with my message. The experience was pleasant if not perfect, and stopping here would leave me feeling pretty good about the exchange.</p><p>And that’s when I remember that I’m talking to the government. The bot makes a jarring shift to interrogating me about my physical location and contact details.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/674/1*IvHmvjG09F0pyjhY8cACWQ.png" /></figure><p>The web form asks for the same details — it lets me see the entire interaction at a glance, asterisks and all, so I know what I’m getting myself into before typing a single character.</p><p>A barrage of questions over chat feels more like an ambush.</p><p>I get it. They need this information to secure the homeland. Unfortunately, the conversational UI becomes a liability because I begin to feel like a stranger wants to know where I live for unknown reasons.</p><p>Of course the government has this information. They won’t give me a driver’s license or the right to vote without it. The chat affordance makes it feel like a direct violation of my privacy and I suspect other users will have a similar response in a range of contexts.</p><p>Privacy aside, these interactions seem even more tedious and unfulfilling than they do in the web form. By a simple count, the majority of my exchanges with the bot come after I said my piece and discharged that energy.</p><h4>Whew, made it. How does it end?</h4><p>The final interaction with the bot is an embedded video of Obama discussing the value he gets from these letters. It’s nice and not too long. Obama’s own natural warmth counterbalances the generic character of the other interactions to some degree.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/675/1*QQMIUA1661R1etDgC85DyQ.png" /></figure><p>He actually reads a few letters aloud that express common public sentiments. If you had written something similar only moments before, I imagine watching Obama speak your thoughts would cause a powerful sense of connection.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/676/1*30cCz9ZKBFMVNmlNl7JcsQ.png" /><figcaption>Obama reads a passage from his diary.</figcaption></figure><p>The final message is a blushing emoji — the cute smiley that expresses warmth and affection.</p><p>I see their intention and yet it feels a bit incongruous after barrage of personal information.</p><h3>Concluding thoughts</h3><p>I expect the bot will simultaneously succeed in its goal and yet fall short of its mission just like the suggestion box at the DMV.</p><p>The conversational interface — I really hesitate to call this a bot — is a clear improvement over the form in discoverability, ease of use on mobile, etc. Submitting my actual comment is now much easier. But I never actually submitted a comment because the subsequent interrogation freaked me out.</p><p>The entire interaction has a three part structure: accept your submission, get your personal info, smother you with affection. I got a jangly, unsettled feeling at each of the transition points that had a net effect of lowering my confidence in the bot.</p><p>What’s really missing is the re-engagement. Every chat is the first chat. No attempt at follow-up. As a bot builder and a citizen, I have to say it’s a missed opportunity to fulfill its greater mission.</p><p>Obama will only read 10 letters from the public every evening, only a few of which originate from the bot. I understand the limitation — he deserves his sleep. Yet a more imaginative approach to the bot would yield so much more value.</p><p>For instance, comments could be aggregated and serve a rough poll of the concerns and sentiments of the public. The White House could then share those results back to me via chat, creating a feedback loop that might keep me engaged with both the bot and the government. Right now, they’re both a bit of stretch to say the least.</p><p>Nevertheless, the White House reached many thousands of people that likely would not have even thought to write Obama a letter through another means. A better title for this post might have been “how to reduce a bot into an empty box and have it succeed anyway.”</p><h4><a href="https://medium.com/ddouble">Follow the ddouble.ai publication</a> to make sure you don’t miss any future Tuesday Bot Teardowns!</h4><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=745441743122" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ddouble/how-the-white-house-made-a-bot-out-of-an-empty-box-745441743122">How the White House made a bot out of an empty box</a> was originally published in <a href="https://medium.com/ddouble">ddouble</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Tuesday Bot Teardown: Growbot]]></title>
            <link>https://medium.com/ddouble/tuesday-bot-teardown-growbot-ff997e785582?source=rss----f293534d269e---4</link>
            <guid isPermaLink="false">https://medium.com/p/ff997e785582</guid>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[bot-teardown]]></category>
            <category><![CDATA[slack]]></category>
            <category><![CDATA[messaging]]></category>
            <dc:creator><![CDATA[Sandi MacPherson]]></dc:creator>
            <pubDate>Tue, 13 Sep 2016 17:41:38 GMT</pubDate>
            <atom:updated>2016-09-13T18:03:01.075Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/638/1*5uD0yaxaNnr9MASLTB-F9Q.jpeg" /></figure><p>The core intention behind <a href="https://medium.com/u/f1c8ab97bfaf">Growbot</a> is to allow the people on your team to share appreciations with each other, supporting each other to do their best work. It’s a light-weight, easy to use bot that’s been funded by some of the <a href="https://techcrunch.com/2016/07/07/bot-or-not/,%20https://techcrunch.com/2016/07/19/slack-now-with-600-apps-on-its-platform-pours-2m-into-11-slackbot-startups-via-its-slack-fund/">best investors</a> in the space.</p><blockquote>“We all work harder when we know our work is meaningful.” — Growbot Co-Founder Joseph Estrada</blockquote><p>To understand Growbot’s success to-date, we looked at the main interactions of the product, where it sits competitively, how the product acquires new users, and the technical stack that Growbot is built on.</p><h3>About Growbot</h3><p>A team admin can invite Growbot to any channel(s). Growbot listens for and encourages praise (looking for keywords ‘props’, ‘kudos’, etc), and keeps a running tally while also giving shoutouts (complete with fun reactions and emojis) when a user in the channel receives props from another user. The bot also delivers a weekly props summary, highlighting the person with the most props as well as the one prop that was most popular.</p><blockquote>Our goal is to build a product that strengthens the habit of appreciation. The more you do it, the better you get at it. The happier you are, the harder you work. —Growbot Co-Founder Joseph Estrada</blockquote><h3>Onboarding</h3><p><strong>Discovery</strong></p><p>Currently, you can find Growbot in the Slack App Directory. You can also get the bot directly via the ‘Add to Slack’ button on the company’s website, at <a href="https://www.growbot.io/">growbot.io</a>. The Growbot team has been putting together some great content marketing pieces too in <a href="https://medium.com/growbot-s-diary">their Medium publication</a>, including <a href="https://medium.com/slack-developer-blog/slack-bot-onboarding-3b4c979de374">this one</a> that was featured on the Slack Platform Blog.</p><p><strong>First experience</strong></p><p>Once growbot has been added to a Slack team, the admin that added the bot receives a DM to add Growbot to a channel. The instructions are clear and to-the-point:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/671/0*BtgbyPSBWaI7OHLQ." /></figure><p>To make sure Growbot puts their best foot forward as they’re meeting teammates for the first time, Growbot first pings the admin directly to ask if an introduction is needed, or if the admin themselves will update the team on how and why to use the new bot:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/662/0*zwVlg_pWJvuLZvQb." /></figure><p>After inviting Growbot to a particular channel and indicating ‘Yes, please introduce yourself!’, Growbot says hello and politely asks anyone in the channel if they’d like more information on how to learn more about using the bot. The use of emoji to express interest is pretty great 👌</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/669/0*BXxxipUnty3g_bFh." /></figure><p>Instead of filling the screen of that channel with added information that’s not needed, this one-to-one approach shuttles the ‘how-to’ information — if it’s requested at all — directly to a DM only to the person who requested more info:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/652/0*y-bqdsM9nlG65TWv." /></figure><p>Again, the instructions are clear and only list what’s needed to start interacting with Growbot. There are links to other content, but the goal here seems to be focused on ensuring that a person can achieve one successful interaction with the product, not understand the complete feature set immediately.</p><p>There was very little friction in adding the bot to Slack, and the onboarding process was super smooth.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/655/1*TsBiJ2hkbtDVndOIsU_yhQ.png" /></figure><h3>😂</h3><h3>Bot UI: Language / Tone / Style</h3><h4>Core interactions</h4><p>Growbot is very friendly and encouraging. To invoke the bot, a user types the words ‘props’, ‘cheers’, or ‘kudos’, followed by an @ mention of the user they’re crediting, followed by an explanation of what they’re giving praise for.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/657/1*oO4qpdgsDBIbLORDvIoLDg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/653/1*xc8ntL1J-ho1IYHvf6t5cA.png" /></figure><p>Growbot automatically adds encouraging reactions (in the 2 examples above, applause and an earth emoji) when it’s invoked, and responds with an equally encouraging phrase. The person receiving the props is listed, along with a tally of how many props they’ve received to-date.</p><p>The reactions from Growbot were great — they were encouraging! Seeing positive messages acknowledging your hard work directly from a teammate is great, but the added fun from Growbot gave an extra dopamine shot to the person receiving the encouragement.</p><p>Growbot is very quick to respond — which, in this case, makes sense. There’s no real thinking that the bot needs to do, and so there’s no need to introduce a false sense of timing around ‘the bot is thinking right now’, as some bots working on more complex problems often insert to imbue their bot with a sense of ‘it’s actually doing work’ to users.</p><p>It seems like, over time and with more data, there’s a potential opportunity for Growbot’s responses to become more contextual based on the input. For example, this piece of encouragement was 100% unrelated to ice cream:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/657/1*favEIDQVIKPVazkcYfDdBw.png" /></figure><p>…but pulling the word ‘diagram’ from the win that was shared could suggest an ‘ 📈 Up and to the right!’ type of response, making it feel a bit more in-tune with the people and the work they’re doing.</p><h4>Language</h4><p>The phrases and language used by Growbot are fairly constrained, focusing just on simple replies to the encouraging phrases it’s listening for. There are lots of emoji used in-line in responses, in the weekly updates, and the reaction emoji are always cute and tie to the messaging in the response itself.</p><p>Another thing that we noticed was that Growbot is very concise, which was greatly appreciated. There were never any points where we were confronted with a confusing wall of text, or felt that the cognitive friction around how to use Growbot was overwhelming.</p><h4>Web experience</h4><p>All of the encouraging reactions that Growbot responds with are also hyperlinks, leading to a web dashboard showing a summary of encouragements received across the team:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1004/1*65kjt9NCFyQgQ3UmEXqALA.png" /></figure><h4>Errors and unanswered messages</h4><p>Some commands that we expected to work didn’t yield any response from Growbot, such as @ mentioning Growbot directly:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/654/1*1L3rppA64_hHAiaOEeXV0A.png" /></figure><p>Similarly, a forgotten @ mention and Growbot doesn’t offer any advice or suggestions on what might be wrong with your attempt at encouraging your teammate:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/654/1*z0afOtBbzjUIGx8sBMuyXQ.png" /></figure><h4>Getting help</h4><p>A simple ‘help’ message in a DM to Growbot pulls up exactly the type of information you’d be looking for if you had questions about how to use the bot:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/655/1*xE-K-l4AmSuSnwdyrSNRxA.png" /></figure><p>One interesting feature of Growbot is the ‘operator’ command. While in a DM with Growbot, if you’re still confused and need extra help, the typing operator will ping a real person from Growbot to join the chat and help understand your problem. It’s very consistent with the mode of interaction, keeping all of the communication between the user and the bot/company in one place.</p><h4>Sharing a summary of activity</h4><p>Every Friday at 10am, the channels where Growbot lives receive a summary of all of the encouraging wins shared by teammates that week. There’s a text summary of the total number of props received, and the most popular win is highlighted and displayed. The most popular reactions to the various props are also highlighted:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/655/1*liudUE1a0F4lIGbgf-fzhQ.png" /></figure><p>There was one other push that I received so far in using Growbot, which was a summary of the past 24h and props I’d given over that span of time:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/565/1*tjC8a8lf7PTMvsb7eo2dJg.png" /></figure><h3>Growth</h3><p>Growbot is very polite when it comes to interacting with other members on your team. It asks for permission before introducing itself, and then further does a quick intro and suggests getting more information about how to use the bot — but only if you really want to! There are no obnoxious asks to ping everyone on your team, nor are there any mentions of spreading Growbot across as many channels as possible. The weekly and 24h summary and messages are the only push messages that Growbot creates and delivers independently, unprompted by any user action.</p><p>The Growbot team is focused on building a delightful experience for end users, not targeting and selling to HR departments — which is the strategy many of their competitors are using. It seems to be working!</p><h3>Technical</h3><p>Growbot is a fairly simple bot atm. It’s very firm in what it will respond to, and doesn’t try to step in where it’s unneeded or where it doesn’t have the context to complete actions that are outside of the stated domain. It appears that the language tools are fairly simple, or at least the design of the bot means that complex NLP/NLU or AI are not needed.</p><p>The paid tier of the product allows teams to customize appreciation words beyond the standard kudos/props/cheers options. Paid accounts also get API access. According a <a href="https://growbot.workable.com/jobs/279717">recent job posting</a> (they are hiring!), the team runs node.js, Redis, and PostgreSQL on AWS.</p><h3>Final thoughts</h3><p>We really like using Growbot, it’s one of the best bots that we’ve tried. Using it on our team of two is fun, and it almost gives us an excuse to be more congratulatory than usual. We can recognize small wins with an easy light-weigh action that might otherwise seem out of place, and therefore not happen at all. If something is fun, more people are willing to do it. And if that fun thing makes for happier employees… then it’s a win-win situation for both employees and their company. Other companies really like it too — <a href="https://medium.com/u/7f210cc75522">Breather</a> wrote an <a href="https://breather.com/blog/culture-club-how-one-little-bot-helps-us-live-our-values-2/">entire post</a> about how Growbot helps solidify their open, friendly company culture.</p><p>The difference between the Free and Paid product makes a lot of sense from a user perspective and feels fair. You get access to additional features if you pay (i.e. custom ‘props’ terms, a dashboard for managers), but can still use a full-fledged, high-quality product without paying.</p><p>We did notice that our use of the ‘props’ terms dropped over time, and there’s some potential for Growbot to have some loops to remind us that the feature exists — perhaps drawing on data around previous frequency of props given, the relationships that are more actively giving props than others, etc. As previously mentioned, the ability to have contextual messages could also make the product a bit more fun… but that’s not an area where it’s currently lacking. It seems like the Growbot team will do a stellar job getting that balance right, as they’re so polite with their existing ways of integrating Growbot into various channels and teams in an unobtrusive way.</p><h4><a href="https://medium.com/ddouble">Follow the ddouble.ai publication</a> to make sure you don’t miss any future teardowns!</h4><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ff997e785582" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ddouble/tuesday-bot-teardown-growbot-ff997e785582">Tuesday Bot Teardown: Growbot</a> was originally published in <a href="https://medium.com/ddouble">ddouble</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Tuesday Bot Teardowns: How the best startups are building amazing bots]]></title>
            <link>https://medium.com/ddouble/tuesday-bot-teardowns-how-the-best-startups-are-building-amazing-bots-59ea347959ab?source=rss----f293534d269e---4</link>
            <guid isPermaLink="false">https://medium.com/p/59ea347959ab</guid>
            <category><![CDATA[tech]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[bot-teardown]]></category>
            <dc:creator><![CDATA[Sandi MacPherson]]></dc:creator>
            <pubDate>Mon, 12 Sep 2016 17:50:08 GMT</pubDate>
            <atom:updated>2016-09-13T02:38:09.199Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*YIcyg53m_I1tDh1J13kZNg.jpeg" /></figure><p>One of the best ways to get a sense for how something works it to take it apart. The understanding that you can get by thinking about the various components of the system and the decisions that were made to create the final product can be illuminating — in a way that simply thinking about an experience with the product can’t quite capture.</p><p>To think more critically and deeply about bots, <a href="https://medium.com/@micaelaneus">Mica</a> and myself will share teardowns of bots that we’ve come across via <a href="https://medium.com/ddouble">this publication</a> (follow it!). We’ll spend time using the bots as intended for at least two weeks, and go through a pre-defined outline to systematically take apart and understand the components of each bot. There will be a new teardown every Tuesday.</p><p>We also know that the overall experience is very important. We won’t only focus on the UI elements, the language choices, or the icons used to represent these human-like entities. We’ll also be adding out our personal experiences to the reviews — how did the bot make us <em>feel</em>? What type of relationship did we develop with the bot, with each other, with other people because of the bot? What changed in our lives with the inclusion of this bot? We’ll also try to put ourselves into the shoes of the people making these bots — why did they make certain decisions? What would we have done differently, and why?</p><h4><strong>We’re excited to get started — the first teardown will be up tomorrow — and hope you’ll </strong><a href="https://medium.com/ddouble"><strong>follow along</strong></a><strong> via the ddouble.ai Medium publication!</strong></h4><p>PS: If you missed it, be sure to check out the <a href="https://medium.com/ddouble/how-to-build-your-best-bot-the-bot-stack-compendium-90a90660167a">Bot Stack Compendium</a> we released last week, to collect all bot-related tools and services</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=59ea347959ab" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ddouble/tuesday-bot-teardowns-how-the-best-startups-are-building-amazing-bots-59ea347959ab">Tuesday Bot Teardowns: How the best startups are building amazing bots</a> was originally published in <a href="https://medium.com/ddouble">ddouble</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>