<?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[Stories by Nancy B. Duan on Medium]]></title>
        <description><![CDATA[Stories by Nancy B. Duan on Medium]]></description>
        <link>https://medium.com/@nancyduan?source=rss-8ad90916b31------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*GzZcuq35z0WVVznh.jpg</url>
            <title>Stories by Nancy B. Duan on Medium</title>
            <link>https://medium.com/@nancyduan?source=rss-8ad90916b31------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 01:11:33 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@nancyduan/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Content Writing for Your Bot]]></title>
            <link>https://medium.com/chat-bots/content-writing-for-your-bot-e0f1dd0320da?source=rss-8ad90916b31------2</link>
            <guid isPermaLink="false">https://medium.com/p/e0f1dd0320da</guid>
            <category><![CDATA[content-design]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[content-strategy]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[conversational-ui]]></category>
            <dc:creator><![CDATA[Nancy B. Duan]]></dc:creator>
            <pubDate>Wed, 15 Mar 2017 03:36:32 GMT</pubDate>
            <atom:updated>2017-03-17T23:37:22.574Z</atom:updated>
            <content:encoded><![CDATA[<p>Bonus part 5 of Designing Conversational UI with Information Architecture</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*oG4QcdS5HiR0Q0P3-_Phqg.jpeg" /></figure><p>Welcome to a bonus part 5 of my series on <a href="https://chatbotslife.com/design-a-conversational-ui-with-information-architecture-5c6ac9576ba3#.844d3bc2i">IA and Conversational UI</a>! I was asked about content writing, and although I no longer drive that portion of my work, I did work in communications so I was happy to write a bit about this.</p><blockquote>There’s an increasing amount of overlap between graphical user interfaces (GUI) and voice user interfaces (VUI), which necessitates a new skillset: user experience professionals with deep understanding of narrative design and conversational design. -<a href="http://www.uxbooth.com/author/kristinabjoran">Kristina Bjoran</a>, UX Booth</blockquote><p>One of the more interesting aspects of content writing for chatbots is the need to move beyond writing for persuasion. Marketing and communications (and some technical writing) is traditionally about conveying a brand voice and ideally soft-selling a product or service. While chatbots can hold this function, content writing for conversational UI should be more akin to UX writing.</p><p>Yes, another UX role.</p><p>Let’s not fall into the trap of defining yet another role for us poor creatives to take on. Let’s just think about this in a more traditional design sense.</p><p><strong>UX writing and IA is about placemaking, sensemaking and wayfinding.</strong></p><p>For the last 4 parts of the series, I talked about sensemaking and a bit about wayfinding. But I didn’t talk much about placemaking.</p><h4>Placemaking is not Spacemaking</h4><p>I have something to confess. I never write about placemaking in IA because I end up in a weird loop. In order to define placemaking, you must define place. But ultimately placemaking is about defining places! Words are hard.</p><p>So here is my attempt at a brief description of placemaking and how it is applicable in the digital realm.</p><p>In (non-information) architecture, we define places <em>literally.</em> But we do this by creating <em>things</em> and <em>spaces</em> that people already understand. Think of them as design patterns.</p><p>In the front of my building is an 8 ft tall piece of wood with a glass pane inserted at above eye-level and a piece of round metal protruding from it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*W-Eg9_742NTqW5UvZlN-lw.jpeg" /><figcaption><a href="https://www.flickr.com/photos/satguru/">David Howard</a> (CC BY 2.0)</figcaption></figure><p>It’s a door! Doors have been around for a long time and although they take many shapes and forms, people almost automatically know what to do with them. They are the entrances and exits to other <em>spaces</em>.</p><p>But what happens when a door no longer looks or acts like a door?</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2Ft47rOdCBwUhWg%2Ftwitter%2Fiframe&amp;src_secure=1&amp;url=http%3A%2F%2Fgiphy.com%2Fgifs%2Ft47rOdCBwUhWg&amp;image=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Ft47rOdCBwUhWg%2Fgiphy.gif&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="326" frameborder="0" scrolling="no"><a href="https://medium.com/media/44b38e345532ff8c5f49466c9567eb95/href">https://medium.com/media/44b38e345532ff8c5f49466c9567eb95/href</a></iframe><p>In an information environment, we make places in a similar manner. People have come to expect certain consistencies in the words and information that we architect. And when our content suddenly shifts from being a more traditional-looking door to a glass pane, we create problems.</p><p>Placemaking ultimately leads to sensemaking and wayfinding.</p><h4>Identifying what place you are architecting</h4><p>Your content will need to have a purpose. And that narrative voice, that content strategy, needs to come from a bot’s purpose. So let’s try defining that first.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*tOWgdTuQNhHgJ9ZzWAL3Gw.png" /><figcaption>From <a href="https://medium.com/u/6220261a62d2">Jorge Arango</a>’s must read: “3 Placemaking Lessons From the Magic Kingdom”</figcaption></figure><p>You can fill in the sentence above to get you started. <em>Here you ask for movie recommendations based on previously-defined preferences. </em>This is fine, but rather product-y. Think grander. <em>Here you come to open up your world and discover new movies.</em></p><p>Remember: You are writing about the <em>place</em> you are making, not the brand or the product. And you are doing it from the point of view of your user, not your bot.</p><h4>What about the content?</h4><p>Great! Now you have a statement that aligns your designers, engineers, marketers and writers. What now?</p><p>You’ll need to define a few other traits before you can jump into the writing. Some of these traits may come from an existing brand or editorial guide. But some of them you will need to define on your own.</p><ol><li><strong>What is the bot’s personality? </strong>Is it jovial? Does it make jokes? Is it a serious bot? What is appropriate for a custom t-shirt maker is different than what is appropriate for a healthcare provider.</li><li><strong>What is the content’s style?</strong> We all know to avoid passive voice when writing. But when you’re crafting conversations, you may want to include some more passive sentences (see what I did there?). It just feels more natural.</li><li><strong>What are the plain-language equivalents of industry jargon?</strong> This doesn’t always come out as a step in content writing, but as someone who comes from some rather difficult-to-communicate fields, I always take the time to do this.</li></ol><p>Now go write a first draft. Outline a back and forth conversation as <em>you</em> understand it.</p><h4>Testing your content</h4><p>Testing conversations has not changed much in years. I was listening to a colleague tell me about a test they ran some 15 years ago for a large telecom giant. They were testing the voice prompts for the phone system and had people sit around a large conference room with scripts and acted out each prompt.</p><p>We can now make and deploy chatbots pretty easily. It’s such a low barrier with chatbots that I recommend deploying content in beta and iterating from there. As someone who is constantly (constantly) refining and editing my writing, I can attest that having real users give me feedback and data has been invaluable to the process.</p><p>So yeah. I didn’t exactly give you a step-by-step guide to writing content. I scoured medium and realized that there are those who have already written much better articles about the subject. So I will point you to them instead.</p><p><strong>Additional Reading:</strong></p><ul><li><a href="https://uxdesign.cc/making-chatbots-talk-writing-conversational-ui-scripts-step-by-step-62622abfb5cf">Making Chatbots Talk — Writing Conversational UI Scripts Step by Step</a></li><li><a href="https://medium.com/dropbox-design/design-words-with-data-fe3c525994e7">Design words with data</a></li><li><a href="http://www.dozuki.com/Tech_Writing">Tech Writing Handbook - Dozuki</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e0f1dd0320da" width="1" height="1" alt=""><hr><p><a href="https://medium.com/chat-bots/content-writing-for-your-bot-e0f1dd0320da">Content Writing for Your Bot</a> was originally published in <a href="https://medium.com/chat-bots">Chatbots Magazine</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Designing Conversational UI with Information Architecture — Part 4]]></title>
            <link>https://medium.com/chat-bots/designing-conversational-ui-with-information-architecture-part-4-1043d21f19cf?source=rss-8ad90916b31------2</link>
            <guid isPermaLink="false">https://medium.com/p/1043d21f19cf</guid>
            <category><![CDATA[design]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[content-design]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Nancy B. Duan]]></dc:creator>
            <pubDate>Wed, 01 Feb 2017 15:20:55 GMT</pubDate>
            <atom:updated>2017-05-17T19:24:52.697Z</atom:updated>
            <content:encoded><![CDATA[<h4>Onboarding without arrows</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*js5DD5ZPoqsPMuYY2oontA.jpeg" /></figure><p>Welcome to part 4 of a <a href="https://chatbotslife.com/design-a-conversational-ui-with-information-architecture-5c6ac9576ba3#.h9v00rqsz">series on designing for conversational UIs</a> using information architecture. This part will be about the art of onboarding when you don’t have a static screen.</p><blockquote>Onboarding is your first date with your user — you don’t want to talk their ear off and you want to be as charming as possible.-<a href="http://www.dtelepathy.com/blog/author/mkdigital-telepathy-com">MK COOK</a></blockquote><p><em>This part will be the least IA focused of the series, but I wanted to draw attention to it.</em></p><p>Why is onboarding so hard? Part of the reason is that designers forget some basic rules while creating onboarding flows. Users of your application will have only so much brainpower/attention span to go through the laborious process of clicking through 7 onboarding screens. Which is why I am a big proponent of empty states that nudge users in the right direction.</p><blockquote>Making a good first impression is important, but true success in your first empty state means driving an action. Make your user want to fill that screen.-<a href="http://blog.invisionapp.com/author/dina-chaiffetz/"><strong><em>Dina Chaiffetz</em></strong></a></blockquote><p>Material design has dedicated a <a href="https://material.io/guidelines/patterns/empty-states.html#empty-states-avoiding-completely-empty-states">portion of their guidelines to empty states</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*qdN-FBO5gThI0ISHhqB3lA.png" /><figcaption>Material Design Guidelines</figcaption></figure><p>A good empty state is clear, concise, and prompts users to take action in a meaningful way.</p><h4>Empty States for Screenless Interfaces? What?</h4><p>Right. So this doesn’t make much literal sense. But I just wanted to start off with empty states, because they are not always seen as an onboarding technique.</p><p>For chatbots, let’s start with how we onboard users at the beginning, before that default menu prompt.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qAFYUVGwOR8xfZiOLGXBPg.png" /></figure><p>Slackbot uses both an empty state and a prompt, knowing that many users will ignore one or both. The first set of helpful tips is designed like an empty state. It tells you what type of question you can ask by giving you an example and then asks you to add a profile picture. The prompt that slackbot then uses reiterates the idea that you can ask questions while starting off the conversation.</p><p>But how would we do this in a non-visual way?</p><h4>Is there anything I can help you find?</h4><p>If you’re in the US, you’ve probably had someone greet you when you enter a large retail store at some point.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9JF2JjcTsEe2QKWDZmpybA.png" /></figure><p>“Welcome to ABCD Super Store, can I help you look for anything?”</p><p>Those of us just browsing may find it a little annoying, but for anyone actually looking for a specific item, greeters can feel immensely helpful in an otherwise overwhelming big box store. Most interactions with your chatbots and voice interfaces will be from users who are specifically looking for something. That may change in the future when we look to AIs more and more for actual conversation, but until then, let’s give our users some help.</p><h4>A word of caution…</h4><p>This should not turn your voice UI into an endless chat with a telephone prompt!</p><p><em>Thank you for calling Fanny’s Repair Services. If you’d like to repair your car, press or say 1. If you’d like to repair your sink, press or say 2. If you’d like to repair your air conditioning, press or say 3. If you’d like to repair your other sink, press or say 4. If you’d like to speak to a human being, yell loudly.</em></p><p>That is an obviously fake example, but not so far off from the truth. The main difference between a phone voice prompt system and your conversational UI should be a human-understandable voice. Not a voice that sounds human. A voice that speaks like a human.</p><p><em>Hey Chuck, thanks for coming to me with your repair problem. Can I help you with anything specific? I can do estimates, give you times I’m available, or diagnose a problem.</em></p><h4>Where IA plays a role</h4><p>It’s time to take that hierarchy and <a href="http://uxmastery.com/testing-information-architecture/">do some usability tests</a> on it.</p><p>I’m not going to detail out the steps of testing an IA (without a screen!) but I will say that this will help you figure out how you can and should onboard your users. The usability test will point out some obvious breakdowns in your work, but for some areas, you may not be able to change enough to vastly improve your content. So instead of scrapping your hierarchy and starting over, use that information for your onboarding prompt.</p><p>Let’s say you’re testing your hierarchy and content model from your movie recommendation bot…</p><p><em>According to your search history, we believe you’d enjoy </em>The Social Network, <em>also by David Fincher. Have you seen it?</em></p><p>…and you discover that people don’t have any idea how you came up with a David Fincher moview as a recommendation. Well, you could clarify in your onboarding by telling users how recommendations are surfaced.</p><p>Or you may discover that usability test participants don’t generally like that recommendations are based on something they can google search themselves (like: David Fincher Movies) and want something more robust. Well, instead of redesigning your entire IA, you may just want to prime users with this information.</p><p><em>Hey Chuck. I’m here to recommend movies based on some information you’ll provide to me through your search history. When you ask for a recommendation, I’ll give you some based on clues that you’ve left, like movie directors, genres, and actors.</em></p><p><strong>Suggested Reading:</strong></p><p><a href="http://uxmastery.com/testing-information-architecture/">How to Test an Information Architecture</a></p><p><a href="http://www.questionablemethods.com/2012/01/test-your-navigation-with-reverse-card.html">Test your navigation with a reverse card sort</a></p><p><em>This was the hardest part of the series to write and still feels slightly disjointed. Please send me a message, leave me some feedback if anything is super unclear!</em></p><p>Although this is a 4-part series, I was asked about content writing for bots in general, so we’ll have a bonus part at some point. Stay tuned!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1043d21f19cf" width="1" height="1" alt=""><hr><p><a href="https://medium.com/chat-bots/designing-conversational-ui-with-information-architecture-part-4-1043d21f19cf">Designing Conversational UI with Information Architecture — Part 4</a> was originally published in <a href="https://medium.com/chat-bots">Chatbots Magazine</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Designing Conversational UI with Information Architecture — Part 3]]></title>
            <link>https://medium.com/chat-bots/designing-conversational-ui-with-information-architecture-part-3-b4746835a517?source=rss-8ad90916b31------2</link>
            <guid isPermaLink="false">https://medium.com/p/b4746835a517</guid>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[content-design]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[design]]></category>
            <dc:creator><![CDATA[Nancy B. Duan]]></dc:creator>
            <pubDate>Tue, 10 Jan 2017 00:46:24 GMT</pubDate>
            <atom:updated>2017-03-17T23:51:06.172Z</atom:updated>
            <content:encoded><![CDATA[<h4>An IA for AI Learning</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2vm9nJuss2cgwE4n_-h4LQ.png" /></figure><p>Welcome to part 3 of a <a href="https://chatbotslife.com/design-a-conversational-ui-with-information-architecture-5c6ac9576ba3#.h9v00rqsz">series on designing for conversational UIs</a> using information architecture. This part will be about information retrieval, natural language processing and taxonomy.</p><p>Wait! Don’t go! I didn’t mean to trick you into reading about taxonomy. But if you’re still here, I will make a case for why its important for AI and information extraction.</p><h4>Taxonomy. Ugh! What Is It Good For?</h4><blockquote><strong>Taxonomy</strong> is the core component of information architecture…Derived from analysis of usage patterns and information flow, taxonomy guides the visual design of information navigation and administration in compliance with standards and guidelines. — <a href="http://www.techrepublic.com/article/understanding-information-taxonomy-helps-build-better-apps/">TechRepublic</a></blockquote><p><a href="https://chatbotsmagazine.com/designing-conversational-ui-with-information-architecture-part-1-e1fafa629031#.te787tyws">In part 1</a>, I talked about content modeling and building a map of inter-related content that is reusable and purposeful. <a href="https://chatbotsmagazine.com/designing-conversational-ui-with-information-architecture-part-2-f636dea4cbd0#.h5n9sk47y">In part 2</a>, I talked briefly about creating a hierarchy of information to identify <em>how</em> your chatbot will call up information.</p><p><strong>Creating a taxonomy will help your bot identify <em>what</em> to call up. Taxonomy is the information of your information.</strong></p><p>When I refer to taxonomies, I am talking about naming and categorization. However, there is not a set definition for this term, as it is sometimes used to refer to IA in general.</p><h4>Informing Your Taxonomy</h4><p>As with a lot of UX design, it can be daunting to know where to start when informing your design. What kind of qualitative or quantitative research should I do?</p><p>There are, of course, many approaches to this part, and I defer to what you enjoy and feel comfortable doing. (I have a slightly related post about <a href="https://medium.com/@nancyduan/what-if-your-user-base-is-everyone-a307ed341d17#.r151a764c">doing research for enterprise software</a>.) Specifically for your taxonomy, you’ll want to at least try to define a few things to help you narrow down your design work. Mostly, this is because taxonomy is linked to your contexts of use. Unlike a content structure, your taxonomy depends on having some constraints in your needs.</p><p>I usually start with data. Shocking, I know. I sift through the analytics of my software product or site and look at <a href="https://www.nngroup.com/articles/ia-warning-signs-analytics/">certain key things</a>.</p><blockquote>Analytics metrics such as pageviews, conversions, entrances, bounce rates, and search query frequency can help identify problems in your category structure. -<a href="https://www.nngroup.com/articles/ia-warning-signs-analytics/">Kathryn Whitenton and Katie Sherwin</a></blockquote><p>To get your thinking-juices flowing, here is a non-comprehensive list of all the ways I’ve started on my taxonomic journey:</p><ol><li>Look at google analytics to identify some usage patterns</li><li>Look at some tags/labels that have been used in my CMS</li><li>(For a customer service page) look at highest hits in knowledge base</li><li>Look at usage data for types of devices</li><li>Crawl twitter for hashtags associated with my content pages</li></ol><p>You get the picture. You can always go a more traditional UX research direction. Interview some users. Create personas. Make a journey map. Identify ideal information flows. But I tend to do qualitative after looking at quantitative.</p><h4>A Taxonomy That Your AI Can Search On</h4><p>Most IA practitioners will tell you that a taxonomic structure will help <em>users</em> find things in your site. Or if they can’t find it, search for it. As I mentioned in part 2, I suggest building a bot that searches on terms and not just one that goes down a decision tree.</p><p>In the case of a converstional UI, a taxonomy will help your <em>bot</em> find and retrieve information.</p><p>Depending on how you’ve built your data model and your database in general, creating a taxonomy will differ greatly. But, think of creating a taxonomy as if your AI is a search engine crawler and your data is individual sites. You’d want to create metadata for the crawlers and tags for your own purposes so you can rearrange information later. Same thing happens when you put together a taxonomy.</p><p>The reason I had you put together a content model before all this was so that you already have a lot of the interdependencies laid out.</p><p>Remember the movie recommendation example I gave? Let’s revisit this:</p><blockquote><strong>User:</strong> <em>What other movies have the same director?</em></blockquote><blockquote><strong>Bot: </strong><em>There are 7 other major motion pictures by David Fincher. Would you like to hear them all or narrow them down by </em>[<em>attribute</em>]<em>.</em></blockquote><p>Identifying each attribute and tagging each specific instance of that attribute with identifiers can make this happen easily. But I also believe that taxonomy will help AIs learn better.</p><h4>Natural Language Processing</h4><p>Looking at chatbot builders, I see a lot of technologists. People who are interested in AI and conversational UI, NLP/NLU and machine learning. I am by no means an expert in any of these fields. The little I do know though is around the idea of syntax and the difficulties of NLP in parsing language.</p><p>Last year google released <a href="https://research.googleblog.com/2016/05/announcing-syntaxnet-worlds-most.html">SyntaxNet</a> to help with NLU. In their paper detailing the project and research, they say in no uncertain terms that “it is critical to tightly <em>integrate learning and search</em> in order to achieve the highest prediction accuracy.” Your chatbot AI — with all its fancy new tags and taxonomic structure — will now be available to search on specific items. And then it will in turn help your NLP engine <em>learn</em>.</p><p>I know I simplified that down quite a lot. This field is extremely complex, just as language is. But I am quite confident that as designers of information, we can give our bots a leg up by helping them learn. And this is just one way I propose we go about it.</p><p><em>As always, please drop me a line with feedback, questions. Anything!</em></p><p>Read Part 4: <a href="https://chatbotsmagazine.com/designing-conversational-ui-with-information-architecture-part-4-1043d21f19cf#.5ngalgv2y">Onboarding without Arrows.</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b4746835a517" width="1" height="1" alt=""><hr><p><a href="https://medium.com/chat-bots/designing-conversational-ui-with-information-architecture-part-3-b4746835a517">Designing Conversational UI with Information Architecture — Part 3</a> was originally published in <a href="https://medium.com/chat-bots">Chatbots Magazine</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Designing Conversational UI with Information Architecture — Part 2]]></title>
            <link>https://medium.com/chat-bots/designing-conversational-ui-with-information-architecture-part-2-f636dea4cbd0?source=rss-8ad90916b31------2</link>
            <guid isPermaLink="false">https://medium.com/p/f636dea4cbd0</guid>
            <category><![CDATA[design]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[content-design]]></category>
            <dc:creator><![CDATA[Nancy B. Duan]]></dc:creator>
            <pubDate>Tue, 27 Dec 2016 17:27:21 GMT</pubDate>
            <atom:updated>2017-03-17T23:48:14.734Z</atom:updated>
            <content:encoded><![CDATA[<h4>Hierarchy: solving for the lack of navigation.</h4><p>Welcome to part 2 of a <a href="https://chatbotslife.com/design-a-conversational-ui-with-information-architecture-5c6ac9576ba3#.h9v00rqsz">series on designing for conversational UIs</a> using information architecture. This part will focus on hierarchy and how we can use it to help people better navigate through the complex question and answer format of a conversational UI.</p><h4>Isn’t hierarchy just another term for sitemap?</h4><p>Technically speaking, they can be. I think of sitemaps as the artifact that comes from identifying a hierarchy. You don’t necessarily have to create a sitemap that looks exactly like a hierarchy or vice versa.</p><p><a href="https://www.nngroup.com/articles/flat-vs-deep-hierarchy/">Neilsen Norman</a> has an older article on navigational hierarchies. How flat or how deep your navigational structure should be will depend on the content. For conversational UI, there is a greater importance placed on not just the structure, but the specificity of it. <a href="https://chatbotsmagazine.com/designing-conversational-ui-with-information-architecture-part-1-e1fafa629031">In part 1</a> I talked briefly about content modeling and identifying an underlying content structure that can be reusable and scalable. Here, you can take your content model and start to organize it in a way that helps people navigate through your chatbot.</p><p>A slight side note, but when you start with the content, you are doing what information architects refer to as a bottom-up architecture. You shouldn’t be worried about doing this in a strict sense. Instead, I want to note that no matter how much you structure your IA from top-down and how much easier this is to build, your users will find ways to think outside that structure. Which is why I believe a more bottom-up approach is better.</p><h4>Begin at the beginning</h4><p>When we build chatbots, we <em>usually</em> build a default menu state. Something that you show users when they fire up the bot and when you can’t understand what they are asking.</p><p>To simplify things, I suggest starting with that default menu state. Ask yourself, <strong><em>what is the primary reason (or reasons) someone is chatting with your bot</em>?</strong></p><p>If you have the time and budget for it, I’d do some user testing here. Do an open card sort on your content model and pull out the top-level buckets for your menu state. Or, find a good space within your current experience that could use a friendly, personified message, like a customer service bot.</p><p><strong>This default menu state should then be treated as the top-level of your hierarchy.</strong></p><h4>Don’t dictate the flow</h4><p>Ok, now you need to figure out what <em>type</em> of hierarchy you want to set. For a chatbot, I’ve tried a few. There are very few instances where a strict hierarchy is a good idea, so I prefer to do one that is somewhere between a multi-dimensional hierarchy and a search engine.</p><p>A strict hierarchy looks much like a decision tree. You can only access certain pieces of information after navigating through a prescribed pathway. We UX designers know the perils of creating a strict user-flow. Happy paths are good, but accommodating non-ideal flows are necessary. In a chatbot, since there is a lack of interactive elements, non-ideal will likely be most of your users.</p><p>Most of the chatbots I’ve interacted with have what seems like a strict, but flat, hierarchy. They probably have something a little more complex, considering their ability to process language, but for the sake of this post, let’s assume this is the case.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5z0U861zWU0HikT_0lpZNQ.png" /></figure><p>Google assistant is the one bot I’ve spoken to that most closely resembles a cross between a multi-dimensional hierarchy and a search engine. The reason seems obvious, since it’s google, but the assistant (she? it? ugh.) tends to give search results when it doesn’t know what I’m <em>strictly</em> asking for.</p><p>This can get really annoying. When I refer to designing a “search” hierarchy, I don’t mean put in a search engine. I mean that your bot should search on the terms presented to you by your user and give the best guess to the answer. <strong>And give the ability to correct itself if a mistake is made.</strong></p><p>The example I pasted here is fine. I asked my google assistant to make me hot cocoa. Obviously, this is an insane request (but probably not for long). So instead of saying “I’m sorry, I don’t know how to do that” the assistant just searches the web for a result. But where this breaks down for me is the prompt that happens afterwards.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4kmAjS_rjcxXS0KIQw3QNA.png" /></figure><p>Why is the number 1 stock prompt after giving me search results is to see search results? And why are 3, and 4 equally irrelevant? I don’t even know what “Sign” means, which means I had to tap it to find out.</p><p>And here is where I make my point.</p><p><strong>Specificity is utterly important in a chatbot hierarchy.</strong></p><p>“Sign” here meant that google assistant would search for the keyword Sign on top of its already searched term (in this case “Hot Chocolate”). How the heck was I supposed to know that?</p><p>If you make things too generic in your hierarchy, you’ll end up with users lost in an endless cycle of call and response.</p><h4><strong>Suggested reading:</strong></h4><ol><li><a href="http://webdesignfromscratch.com/website-architecture/ia-models/">Web IA Models</a></li><li><a href="https://www.asis.org/Bulletin/Aug-10/AugSep10_Brown.pdf">Eight Principles of Information Architecture</a></li><li><a href="http://www.howtomakesenseofanymess.com/chapter/2/page/28/start-with-why/">How to Make Sense of Any Mess: Start with Why</a></li><li><a href="https://www.smashingmagazine.com/2013/12/efficiently-simplifying-navigation-information-architecture/">Efficiently Simplifying Navigation</a></li></ol><p>As always, please let me know if you have any thoughts on this or any of the posts in this series. I’d love to hear feedback!</p><p><a href="https://chatbotsmagazine.com/designing-conversational-ui-with-information-architecture-part-3-b4746835a517#.megg7wxna"><strong>Read part 3 here: An AI for IA learning.</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f636dea4cbd0" width="1" height="1" alt=""><hr><p><a href="https://medium.com/chat-bots/designing-conversational-ui-with-information-architecture-part-2-f636dea4cbd0">Designing Conversational UI with Information Architecture — Part 2</a> was originally published in <a href="https://medium.com/chat-bots">Chatbots Magazine</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Designing Conversational UI with Information Architecture — Part 1]]></title>
            <link>https://medium.com/chat-bots/designing-conversational-ui-with-information-architecture-part-1-e1fafa629031?source=rss-8ad90916b31------2</link>
            <guid isPermaLink="false">https://medium.com/p/e1fafa629031</guid>
            <category><![CDATA[content-strategy]]></category>
            <category><![CDATA[content-design]]></category>
            <category><![CDATA[conversational-ui]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[chatbots]]></category>
            <dc:creator><![CDATA[Nancy B. Duan]]></dc:creator>
            <pubDate>Tue, 06 Dec 2016 20:02:51 GMT</pubDate>
            <atom:updated>2017-08-19T20:41:08.167Z</atom:updated>
            <content:encoded><![CDATA[<h4>Content modeling to inform your data model.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Si6lQpi25kDdADfM_eoV-A.jpeg" /></figure><p>Welcome to part 1 of a 4-part series I’m writing about conversational UI and information architecture. I’m starting off with content modeling — what is sometimes seen as a <a href="http://www.iasummit.org/tag/content-modeling/">humble art</a><em>.</em></p><p>It’s not sexy. It may actually be the most tedious of IA jobs, but a very good place to start. It is also, I caution, not a simple task.</p><p><em>This post will not be a step-by-step guide on building a content model for your chatbot. It is, however, designed to get you started down that path.</em></p><blockquote>Content modeling is hard because there are many different ways to achieve that same result, but one way will be more elegant than another. You just have to find that way. — <a href="https://plus.google.com/112663052810877331575?rel=author">Cleve Gibbon</a></blockquote><h4>What Is a Content Model?</h4><p>A content model is not a decision tree. It is not a menu. It is not a hierarchy. It is not a set of editorial guidelines. It is not married to a single web page or device or even user type. Content models are <em>context-independent</em>. This allows us to create scalable, future-proof content.</p><p><strong>A content model is the underlying map of related content types that can help create reusable content sourced from a single place.</strong></p><h4>Start with a Content Audit</h4><p>How you do your content audit will be determined by what you want to do with your content model. If you’re looking to completely overhaul your organization’s content structure, you may want a comprehensive approach. But if you’re just looking to start somewhere so you can design a better chatbot, you can go smaller.</p><p>I am a firm believer that you don’t need to know <em>all</em> your content in order to do a content model. What you do need, though, is a large cross-section of content to help you map it all out.</p><h4>Build a First Model</h4><p>If you were designing a chatbot that is giving movie recommendations, you’d want to get a lay of the land. The content audit should help you identify all the attributes of a <em>movie</em> and all the traits of a <em>recommendation. </em>And then finally you would need to identify what qualities/attributes are inter-related and can be used to put together a <em>movie recommendation.</em></p><p>Here is where a lot of information architects will tell you that the level of granularity of your model will depend upon you. But if you’re doing a content model for the first time, I would attempt to use as much of the content you’ve collected from your content audit as possible, and whittle it down from there.</p><h4>Iterate</h4><p>Rework your content model. Add attributes. Remove inter-dependencies. Group them together to form something new. Whatever it is, do something different to see if something else comes out.</p><p><em>Featured CBM: </em><a href="https://chatbotsmagazine.com/19-best-practices-for-building-chatbots-3c46274501b2"><em>19 Best UX Practices for Building Chatbots</em></a></p><h4>Why a Content Model Will Help with Your Conversational UI</h4><p>At this point you may be wondering why this would help with your chatbot.</p><p>That movie recommendation content model now has all your inter-dependencies and attributes laid out.</p><p>So if a user asks “What other movies have the same director?” your bot can search that attribute [<em>director</em>]. And because you’ve modeled the idea of a recommendation as well, you can not only give one answer, you can give several answers, or something else entirely.</p><p>You may decide a recommendation = <em>There are 7 other major motion pictures by David Fincher. Would you like to hear them all or narrow them down by </em>[<em>attribute</em>]<em>.</em></p><p>You may decide a recommendation = <em>According to your search history, we believe you’d enjoy </em>The Social Network, <em>also by David Fincher. Have you seen it?</em></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2F2XflxztXjThQLL6AhHO%2Ftwitter%2Fiframe&amp;src_secure=1&amp;url=http%3A%2F%2Fgiphy.com%2Fgifs%2Fgiffffr-2XflxztXjThQLL6AhHO&amp;image=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2F2XflxztXjThQLL6AhHO%2Fgiphy.gif&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="179" frameborder="0" scrolling="no"><a href="https://medium.com/media/99062dc0d6f058845288873ca5e745ea/href">https://medium.com/media/99062dc0d6f058845288873ca5e745ea/href</a></iframe><p>This type of content structure helps you build out infinitely more scalable answers, whereas a decision tree can usually only work, well, as a tree.</p><p>And because a content model is context-independent, you can use the same model and structure for future products.</p><h4>How Is This Different Than a Data Model?</h4><p>If you ask developers, it isn’t. A content model can be a mirror image of your data model. But if needed to start somewhere, identifying a content structure should be first, before creating a data model from there.</p><p>As with every system and product, there’s bound to be items outside your content model or structure. In a conversational UI, edge-case scenarios may be the majority of them. <a href="https://chatbotsmagazine.com/designing-conversational-ui-with-information-architecture-part-2-f636dea4cbd0#.lm9dqiq0n">In a future post, I’ll be talking about hierarchy and giving user options without a menu.</a> Stay tuned!</p><h4>Suggested Reading</h4><p>Because this is not a step-by-step process, I recommend the following if you want more details about content audits and modeling:</p><ol><li><a href="http://uxmastery.com/how-to-conduct-a-content-audit/">How to Conduct A Content Audit</a></li><li><a href="https://medium.com/shopify-ux/6-steps-to-product-content-audit-perfection-fc89d3b637f4">6 steps to product content audit perfection</a></li><li><a href="http://www.clevegibbon.com/content-modeling/content-models/conceptual-content-model/">The Conceptual Content Model</a></li><li><a href="http://adaptivepath.org/ideas/figuring-out-content-strategy/">Figuring Out Content Strategy</a></li></ol><p><a href="https://chatbotsmagazine.com/designing-conversational-ui-with-information-architecture-part-2-f636dea4cbd0"><strong><em>Read part 2 here, Hierarchy: solving for the lack of navigation.</em></strong></a></p><p><em>In February I’m putting together a workshop for </em><a href="http://www.2016.worldiaday.org/location/pittsburgh-pennsylvania-united-states"><em>World IA Day - Pittsburgh</em></a><em>. Come and talk content models with us!</em></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fupscri.be%2F58521e%3Fas_embed%3Dtrue&amp;dntp=1&amp;url=https%3A%2F%2Fupscri.be%2F58521e%2F&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=upscri" width="800" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/c90883b0516ddbe5c59960c12a04bc64/href">https://medium.com/media/c90883b0516ddbe5c59960c12a04bc64/href</a></iframe><figure><img alt="" src="https://cdn-images-1.medium.com/max/705/1*mToBIVQSczojw0mSn2h5Sw.gif" /><figcaption>👏👏Clap below to recommend this article to others👏👏</figcaption></figure><figure><a href="https://docs.google.com/forms/d/e/1FAIpQLSfJDqeR7zbRBfQvx6mouQAhjK4lrCy3z9Z6Ppxln7f1xi12dg/viewform?c=0&amp;w=1"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*rRjUc6JW61AP1UWsa-yItQ.png" /></a></figure><figure><a href="https://chatbotsmagazine.com/be-featured-in-front-of-thousands-of-people-interested-in-bots-e7040c4080df#.hswzhzpp9"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*OGAA3ICxv3ds-RK2Sq4oxA.png" /></a></figure><figure><a href="https://chatbotsmagazine.com/want-to-receive-the-best-chatbot-related-content-in-your-medium-feed-20a9c11f11b6#.1t1yhwcyz"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*1XecGp_cGZ1wGLAldRHLWQ.png" /></a></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e1fafa629031" width="1" height="1" alt=""><hr><p><a href="https://medium.com/chat-bots/designing-conversational-ui-with-information-architecture-part-1-e1fafa629031">Designing Conversational UI with Information Architecture — Part 1</a> was originally published in <a href="https://medium.com/chat-bots">Chatbots Magazine</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Do we need more guides for design methods?]]></title>
            <link>https://medium.com/@nancyduan/do-we-need-more-guides-for-design-methods-14cba0ba97f5?source=rss-8ad90916b31------2</link>
            <guid isPermaLink="false">https://medium.com/p/14cba0ba97f5</guid>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[design-thinking]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[user-experience]]></category>
            <dc:creator><![CDATA[Nancy B. Duan]]></dc:creator>
            <pubDate>Tue, 06 Dec 2016 13:53:08 GMT</pubDate>
            <atom:updated>2016-12-06T13:53:08.073Z</atom:updated>
            <content:encoded><![CDATA[<h4>We need fewer how-tos and more value-focused discussions.</h4><p>While writing my <a href="https://chatbotslife.com/design-a-conversational-ui-with-information-architecture-5c6ac9576ba3#.1ni4f1f8y">upcoming series of conversational UI and information architecture</a>, I started re-reading a few books that have influenced me greatly over the last decade. One such book is <a href="https://monoskop.org/images/f/ff/Alexander_Christopher_Notes_on_the_Synthesis_of_Form.pdf"><em>Notes on the Synthesis of Form</em></a> by Christopher Alexander.</p><p>This is a book by an architect that ended up influencing the development of object-oriented programming. It crosses boundaries in the way that it thinks.</p><p>In UX, we’re taught step-by-step methodologies and told to create a very specific artifact from these methods. Methods that have been defined and refined by practitioners.</p><p>Need to identify and understand your stakeholders? Create a stakeholder map! Need to understand the linear progression of users and their actions? Create a user journey map! Want to design a new service experience? Look at some gaps in your service blueprint!</p><p>These methods are great. We all use them. They help us abstract complex interactions, people, and ideas into understandable and digestable information. That’s what we need as designers; Bring clarity to complexity.</p><p>Alexander’s book is roughly 130 pages of a step-by-step method of identifying and creating a pattern based (loosely) on mathematical reasoning. While writing the book, Alexander focused on this part, but ultimately in hindsight, realized the importance of the abstracted information as a way to communicate about the design problem.</p><blockquote>In this book I presented the diagrams as the end results of a long process; I put the accent on the process, and gave the diagrams themselves only a few pages of discussion. But once the book was finished, and I began to explore the process which I had described, I found that the diagrams themselves had immense power, and that, in fact, most of the power of what I had written lay in the power of these diagrams. — <strong>Preface to the paperback edition</strong> (written 10 years after publishing)</blockquote><p>Also in hindsight Alexander recognizes the importance of not just taking his methodology at face value.</p><blockquote>No one will become a better designer by blindly following this method, or indeed by following any method blindly.</blockquote><p>What we do as designers, he argues, is about bringing some kind of order or logic to our work. It is as much about convincing others of the importance of what we are doing as it is about understanding the current state.</p><blockquote>A logical picture is easier to criticize than a vague picture since the assumptions it is based on are brought out into the open.</blockquote><p>So I took this to heart. I realized I could give people a framework for doing something, but not a step by step process. And I’d focus a little more on the value of the artifact and not just the creation of it.</p><p>What do you think? Do you find guides helpful?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=14cba0ba97f5" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Design a Conversational UI with Information Architecture]]></title>
            <link>https://blog.chatbotslife.com/design-a-conversational-ui-with-information-architecture-5c6ac9576ba3?source=rss-8ad90916b31------2</link>
            <guid isPermaLink="false">https://medium.com/p/5c6ac9576ba3</guid>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[conversational-ui]]></category>
            <category><![CDATA[bots]]></category>
            <dc:creator><![CDATA[Nancy B. Duan]]></dc:creator>
            <pubDate>Mon, 28 Nov 2016 17:35:38 GMT</pubDate>
            <atom:updated>2017-02-07T02:49:55.797Z</atom:updated>
            <content:encoded><![CDATA[<h4>Excuse me, what can I ask you?</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hI3qIRoKf3VDEoZgkpvAWA.jpeg" /></figure><p>Ya’ll. Information architecture isn’t sexy. We don’t talk about it like we talk about microinteractions. Or CSS animations. But IA is the backbone of good, usable content and design.</p><p>It is more important than ever to have a solid, detailed IA when you’re building a conversational UI. You’ve stripped the interface of some of its <em>sexier</em> (IE visual) components and now you’re left with words. And these words need to stand alone.</p><p>I recently asked designers to <a href="https://medium.com/@nancyduan/chatbots-need-to-be-designed-not-just-built-87a398750e66#.rb3t7nuoy">start designing chatbots</a>. Right. But how?</p><p>As UX designers we know that a well-designed product helps people navigate, solve problems, identify errors, and do the thing they intended to do. IA does these same things, at a content level.</p><blockquote>If you’ve ever tried to use something and thought, “where am I supposed to go next?” or “this doesn’t make any sense,” you are encountering an issue with an information architecture. <a href="http://www.iainstitute.org/what-is-ia">— Information Architecture Institute</a></blockquote><p>We’ve been using IA since we started giving people information. Architecture (the non-information kind) is one of the oldest professions in the world. But I’d argue information architecture is old as well. Architects built ancient stores; Information architects put signs above those ancient doorways.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*ag70hjwSjE1O0oI3zrzp_w.jpeg" /><figcaption><strong>A dairy shop sign discovered in Pompeii.</strong> <em>Photo by </em><a href="https://www.flickr.com/photos/argyle64/"><em>David Sivyer</em></a><em> CC BY-SA 2.0</em></figcaption></figure><p>Over the next few weeks, I’ll be writing a series on designing conversational UIs using some tried and true IA methodologies. And some I’ve come up with that are hybrids between traditional IA and UX.</p><ol><li><a href="https://chatbotsmagazine.com/designing-conversational-ui-with-information-architecture-part-1-e1fafa629031#.pnrg4ul5p">Content modeling to inform your data model.</a></li><li><a href="https://chatbotsmagazine.com/designing-conversational-ui-with-information-architecture-part-2-f636dea4cbd0#.rnnwk4qa5">Hierarchy: solving for the lack of navigation.</a></li><li><a href="https://chatbotsmagazine.com/designing-conversational-ui-with-information-architecture-part-3-b4746835a517#.ovptzkuat">Designing a system-focused IA that helps users and the AI learn.</a></li><li><a href="https://chatbotsmagazine.com/designing-conversational-ui-with-information-architecture-part-4-1043d21f19cf">Onboarding without arrows</a>.</li></ol><p>If there’s anything you’d like to see that I may not be covering, drop me a line.</p><p><em>In February I’m taking part in </em><a href="http://www.2016.worldiaday.org/location/pittsburgh-pennsylvania-united-states"><em>World IA Day here in Pittsburgh</em></a><em>. We’ve designed a workshop to help people build some information models. </em><a href="http://www.2016.worldiaday.org/"><em>You can check out events local to you</em></a><em>.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/725/1*bQlRSzFHJEmF4Q7PyrLgng.gif" /></figure><figure><a href="https://docs.google.com/forms/d/e/1FAIpQLSeWqtQR0M5J8kRY9fs18qAu746nJJvFyfjc7I2r_El_nuvCBw/viewform?c=0&amp;w=1"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*6XUspT9JOSq0w0Fi35HIaA.png" /></a></figure><figure><a href="https://docs.google.com/forms/d/e/1FAIpQLSfg2do3yC1zW5hFfwS_ute-ZHBHVslrDGJ08Lh3WlfjZQc1uw/viewform?c=0&amp;w=1"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*TISPT5A-k_hgnR-9ibwjhQ.png" /></a></figure><figure><a href="https://docs.google.com/forms/d/e/1FAIpQLSdtGYqBkbZaI0riHgt8SYZkrBTIjrS-Ka5bf9GjIVjO5QLWRw/viewform?c=0&amp;w=1"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*D0Jf3dI6ZThtqcfwDYY7mg.png" /></a></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5c6ac9576ba3" width="1" height="1" alt=""><hr><p><a href="https://blog.chatbotslife.com/design-a-conversational-ui-with-information-architecture-5c6ac9576ba3">Design a Conversational UI with Information Architecture</a> was originally published in <a href="https://blog.chatbotslife.com/">Chatbots Life</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Chatbots need to be designed, not just built.]]></title>
            <link>https://blog.chatbotslife.com/chatbots-need-to-be-designed-not-just-built-87a398750e66?source=rss-8ad90916b31------2</link>
            <guid isPermaLink="false">https://medium.com/p/87a398750e66</guid>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[ux-design]]></category>
            <dc:creator><![CDATA[Nancy B. Duan]]></dc:creator>
            <pubDate>Fri, 25 Nov 2016 17:20:55 GMT</pubDate>
            <atom:updated>2016-12-01T19:24:55.622Z</atom:updated>
            <content:encoded><![CDATA[<h4>A lack of a full interface means the experience has to be well-tailored.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/469/1*6ah1SDSwlm-a9buzy_HlBA.png" /><figcaption>And then the “conversation” stalled.</figcaption></figure><p>Talking to chatbots feels like a lot of work. Either you get bombarded with questions so they can go down their little decision tree or you are presented with a bunch of buttons. What exactly is different about this than, say, a menu? Or a form?</p><p>The chatbot above is one of the better ones. It’s built as a tool to help you converse better in English. I can spend the entire time giving it one word answers and it still valiantly plods along like that guy in your gym class who is drumming up the nerve to ask for your number.</p><p>And then suddenly the bot responds back to me and doesn’t ask a question. And I freeze. What can I say to it? Can I ask?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/530/1*dC7hZF6LGpiTGoffffxsKg.png" /></figure><p>I impulsively replied to that with “YIKES.”</p><p>What if someone you just met talked to you like this? Asking question after question and then suddenly going silent when they ran out of questions to ask? And then when you asked too many questions they lashed out at you?</p><p><strong>Beyond not simulating a real-world conversation, would you continue using an app or site that talked to you like that? What happened to the user is always right?</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/509/1*weNeMartt8mLF2BCn43LlA.png" /><figcaption>Is it contagious?</figcaption></figure><p><strong>Designers.</strong> We need to be more involved in these chatbot builds. I see quite a few of them coming out of marketing departments. Or being built by developers without UX input. Which is fine. This is how we started out when we first got access to an appstore. But we know better now.</p><p>Let us stop making technology for the sake of technology. Chatbots, being devoid of a lot of interactive, screen-based components, need to have their experience very well-designed. And prototyped. And tested.</p><p>Just ask this guy:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/481/1*3vpSru0n5NkmU0XOCTHJsg.png" /></figure><p>Better yet, don’t.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/725/1*bQlRSzFHJEmF4Q7PyrLgng.gif" /></figure><figure><a href="https://docs.google.com/forms/d/e/1FAIpQLSeWqtQR0M5J8kRY9fs18qAu746nJJvFyfjc7I2r_El_nuvCBw/viewform?c=0&amp;w=1"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*6XUspT9JOSq0w0Fi35HIaA.png" /></a></figure><figure><a href="https://docs.google.com/forms/d/e/1FAIpQLSfg2do3yC1zW5hFfwS_ute-ZHBHVslrDGJ08Lh3WlfjZQc1uw/viewform?c=0&amp;w=1"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*TISPT5A-k_hgnR-9ibwjhQ.png" /></a></figure><figure><a href="https://chatbotslife.com/how-to-get-a-free-chatbot-b1fb9dfe109#.z9dtp2sy0"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*D0Jf3dI6ZThtqcfwDYY7mg.png" /></a></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=87a398750e66" width="1" height="1" alt=""><hr><p><a href="https://blog.chatbotslife.com/chatbots-need-to-be-designed-not-just-built-87a398750e66">Chatbots need to be designed, not just built.</a> was originally published in <a href="https://blog.chatbotslife.com/">Chatbots Life</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What if your user base is everyone?]]></title>
            <link>https://medium.com/@nancyduan/what-if-your-user-base-is-everyone-a307ed341d17?source=rss-8ad90916b31------2</link>
            <guid isPermaLink="false">https://medium.com/p/a307ed341d17</guid>
            <category><![CDATA[healthcare]]></category>
            <category><![CDATA[user-research]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[design]]></category>
            <dc:creator><![CDATA[Nancy B. Duan]]></dc:creator>
            <pubDate>Tue, 22 Nov 2016 16:26:38 GMT</pubDate>
            <atom:updated>2016-11-23T15:52:57.281Z</atom:updated>
            <content:encoded><![CDATA[<h4>Designers everywhere: We can’t design for everyone.</h4><p>Designer over here: But what if my product really is <em>for everyone?</em></p><p>You may think that is hyperbole up there, but there are many of us that design complex, multi-modal enterprise systems. But, to be fair, that user base isn’t <em>everyone. </em>Just anyone who may need to get paid from our HR system, or check their bank balance, or file a claim with their insurance.</p><p>I had a recent conversation that went something like this:</p><p><strong>Friend in Healthcare Insurance UX:</strong> I’ve given up on understanding my users. We keep trying to narrow down who our user is, but nothing makes sense. Even if I were to say they all live in a certain region, what difference does that make? Someone could move to our area tomorrow and not have any of the shared interest or needs. What’s the point?</p><p><strong>Me:</strong> Well…I’ve thought a lot about this…</p><p><strong><em>Wait! Let me backtrack a little.</em></strong></p><p>Those of you who know me know I come from healthcare. Mental health care, public health, environmental health, health systems, women’s health. I’ve worked on all these sectors. And so, yes. I <em>have</em> thought a lot about this.</p><p><strong><em>Ok back to the story.</em></strong></p><p><strong>Me:</strong> Personas are about giving your design some constraints. So if you can’t build a simple persona for your user you have to identify your constraints in some other manner.</p><p>I wasn’t nearly that eloquent. But I get to edit my own words.</p><p>All we do as designers, as researchers, are to help solidify some of the inherent constraints in what we’re making. If your system is meant for everyone over the age of 18 with access to a computer and wants relevant health information, a persona isn’t going to help you.</p><h4><strong>Here are some other ways you can identify your own constraints:</strong></h4><ol><li>The Scientific Method: Hypothesis — Test — Refined Hypothesis</li><li>Start with Data: Beyond the market analysis</li><li>Assumption-based Planning: Identifying implicit assumptions first</li><li>Empathize Anyway: Using empathy as a constraint</li></ol><h4>The Scientific Method: Hypothesis — Test — Refined Hypothesis</h4><p>Someone way more eloquent than I (<a href="https://medium.com/u/6df50a3b3770">Micah Sivitz</a>) has written about this very topic. <a href="https://medium.com/shyp-design/a-designer-s-scientific-method-12671b41efb7#.6v8rlk16r">You should read it!</a></p><p>But it boils down to this: Turn a question (“Why is no one using our mobile app for insurance information?”) into a hypothesis (“Our users don’t want to access sensitive information over their phone.”). Then develop a way to test this hypothesis. In this case, you may interview a few users who have the mobile app but do not access their insurance information there. Or build a low-fidelity prototype that pushes the boundary of comfort.</p><h4>Start with Data: Beyond the market analysis</h4><p>No, I don’t mean stare at that dashboard until you have a eureka moment. I’m referring to something that is sometimes called Contextual Analytics or Observational Analytics. Which is just a fancy way of saying <em>identify the why behind the numbers and solve for that</em>.</p><blockquote><a href="http://www.uxbooth.com/articles/putting-big-data-in-context/">Contextual analytics is what allows businesses to trace patterns and detect trends like we did. — <strong>Bartosz Mozyrko</strong></a></blockquote><p>Bartosz gives a great example of using data to hone in on a potential problem to solve, but only using the data to identify <em>what</em> needs a solution, not <em>how</em> to solve the problem.</p><h4>Assumption-based Planning: Identifying implicit assumptions first</h4><p>The step-by-step of this method is very similar to the scientific method, but goes a bit further to identify the implicit and explicit assumptions. You should take the time to identify the implicit assumptions of a product that your explicit assumptions are most likely relying on.</p><p><strong>For example:</strong></p><p>An explicit assumption could be <em>Users prefer to call us when they have questions instead of using the knowledge base.</em></p><p>The implicit assumptions behind that are:</p><ol><li><em>Users find the knowledge base useful.</em></li><li><em>Users know where to go for help when they have questions.</em></li><li><em>Users don’t have any questions that are not answered by our knowledge base.</em></li></ol><p>Once the implicit assumptions are listed out, it is usually clear that the designer may need to delve deeper into a problem before knowing where to focus. If #1 and #2 are true, but #3 is false, than you can develop a test that identifies where the breakdown is happening.</p><h4>Empathize Anyway: Using empathy as a constraint</h4><p>I couldn’t get through writing a post about design without at least mentioning empathy. We designers use it a lot to build ourselves up into user advocates, but how do we use empathy to help us identify what it is to focus on?</p><p>A persona is not just about constraints, it is about putting a name and a face to the problem at hand. So if you can’t put together a persona, why not actually go visit some people with real names and real faces?</p><p><strong>Me:</strong> I’d go to a specialist’s office and ask to sit in the waiting room. I’d talk to people who were waiting. Patients, their caretakers, the receptionist. I wouldn’t ask questions, just talk. Observe. Feel something. Don’t give up.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a307ed341d17" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What I learned about learning,]]></title>
            <link>https://medium.com/@nancyduan/what-i-learned-about-learning-172e7891c99b?source=rss-8ad90916b31------2</link>
            <guid isPermaLink="false">https://medium.com/p/172e7891c99b</guid>
            <category><![CDATA[life-lessons]]></category>
            <category><![CDATA[education]]></category>
            <category><![CDATA[design]]></category>
            <dc:creator><![CDATA[Nancy B. Duan]]></dc:creator>
            <pubDate>Tue, 17 May 2016 17:35:35 GMT</pubDate>
            <atom:updated>2016-05-17T17:35:35.847Z</atom:updated>
            <content:encoded><![CDATA[<h4>or Make, Rattle and Roll</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nXCfRLczuyMZkSJChQIojA.jpeg" /></figure><p>That title is a nod to Rocky and Bullwinkle. Which is my way of segueing into this sentence: <strong>I went back to school at age 29</strong>. I entered into a masters program being well-aware that I would turn thirty while waist-deep in final exams, throwing out outdated references like the one above.</p><p>I was prepared to have more culturally relevant conversations with my professors and not my peers. I was ready to learn from the young kids, some who were busy being born while I was trying to sneak a copy of <em>Dookie</em> into my room unnoticed. I was ready to enter back into the world of learning as I knew it; one where I spent countless hours talking to my peers, passing ideas through anyone who’d give me a minute, critiquing others work, pushing each other far beyond our comfort zones.</p><p>What I got instead was something I was completely unprepared for. I was told over and over again that I was wrong. <em>No, Nancy. Learning is about doing things right. Learning is about avoiding mistakes. Learning is about compromising for a good grade. </em>Here I was, full of excitement and anticipation for the program that I spent years looking for. And I get there and my peers tell me I’ve been doing it wrong for the last decade.</p><p>In 2009, I made a really difficult decision. I sat in a drafty apartment in Somerville, Massachusetts that I shared with 3 other grad students and stared at the makeshift wallpaper behind my desk. My project, a tessellation of triangles that I forced to look like a building, had once again been called out by critics as cerebral and ‘strange.’ I was told that I was making objects, not buildings, and I had convinced myself that I may never be an architect. Unfortunately, I was in a masters of architecture program. A few days later I filed my papers to postpone my upcoming semester. Within a month I had officially withdrawn from school.</p><p>I never became an architect, but my failure stuck with me as I moved through jobs. I bounced around the architectural world, working in every possible discipline that surrounded it from construction to planning to writing about design. None of it stuck, and so I just started trying things. I made print materials and advertisements. I took over social media accounts and taught myself the entire Adobe creative suite. I freelanced and had real jobs and eventually became a director of communications for a nonprofit. I had no reason to accomplish these things. Lots of my projects and jobs were gotten by sheer willpower or luck (or bad luck). I felt like I was a fraud.</p><p>Which brought me back to grad school. I wanted to stop feeling like a fraud. I wanted to be surrounded by other creative types again, learning from the world that enticed me in undergrad. Emerge myself in the discipline without real-world consequences. School was the ultimate playground for me. I could be cerebral and strange again and not take into account the company’s bottom line. But I also knew that after working for 7 years, I could also bring in my real-world knowledge and help inform my designs so they can be useful.</p><p>The program that I chose was not architecture. I couldn’t justify another 3 years of grad school. And I still didn’t want to be an architect. I ended up choosing ux design. Which, if you think about it, is architecture for non-buildings.</p><p>The first semester of school I was perpetually confused. I spent much of the time feeling like an outsider, someone existing on an entirely different planet than my peers. But these feelings were not due to my age. Not due to social constructs or prior knowledge. There was a fundamental difference between how we viewed the world.</p><h3>Make</h3><p>I am a designer. I am confident about that now. I wasn’t when I entered school. I never felt I was good enough for the moniker. If you go back and look at my bios you’ll see I always qualified it. The closest I got was saying ‘I sometimes design things.’</p><p>This program taught me that what I thought were weaknesses were actually strengths. I never felt like a good student. I had always chosen classes based on fascination or waning interests. In undergrad I optioned for courses that were taught by Pulitzer-prize winners and heads of global organizations. Not because they fell under my architecture major, but because, well, why not? My grades suffered from this outlook. Taking a course about weapons of mass destruction without a science background meant a struggle to pass. But I can still tell you at a high level why sticking two pieces of uranium together will make a big boom.</p><p>I took this approach in this program as well. I took a grad-level Behavioral Economics course knowing full well I’ve never taken a social science class in my life. I wrote papers for the first time in 6 years. I struggled and pushed and studied. Importantly, I learned so much.</p><p>In my second semester I wanted to better understand the struggles that my peers were going through. I asked the director of our program if I could do an independent study project under her. I’ve just finished writing up the findings from the semester, and hopefully someday I can post them here. But I wanted to talk about them generally too.</p><p><strong>Making is about creativity. But creativity is about curiosity.</strong></p><blockquote>“I don’t believe in learning from other peoples pictures. I think you should learn from your own interior vision of things and discover, as I say, Innocently, as though there had never been anybody.” <br>― <a href="http://www.goodreads.com/author/show/67899.Orson_Welles"><strong>Orson Welles</strong></a></blockquote><p>There is no right or wrong way to be curious. There is no right or wrong way to be creative. But somehow when it gets to the making part, my peers felt strongly that there were right and wrong ways to make things.</p><p>When creativity must manifest itself and be critiqued and iterated upon, people want help. And they seek help in terms of guidelines and rules. That’s dangerous.</p><h3>Rattle</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/400/1*52Ufq6ctAIABr1QQw_9M2w.gif" /></figure><p>Critique is hard. Hard to give and hard to swallow. But critique is the backbone of design. I can express myself endlessly in an medium and ask the world to accept my viewpoints, but ultimately without critique and iteration, I’d be making art.</p><p>For anyone who has never sat through a critique, it can be brutal. While my current program never devolved into back-breaking slamfests, critique — in whatever variety — is ego-bruising.</p><p>As I mentioned, I came from architecture. One critic I once had told a fellow classmate to ‘draw until the ugly gene came out.’ Another refused to critique my project because I had not done what they had expected with it. I once used a colored ink in a paper model and my critic ‘couldn’t even look at it.’ He preferred monochrome. I’ve also heard numerous stories of architecture school from the late nineties. Critics would literally set models on fire. Take bright red markers and cross out entire drawings in front of peers and industry professionals.</p><p>Critique is extremely arbitrary. I know many many designers who will fight me on this, but ultimately, I don’t believe that’s a bad thing. It teaches students and professionals to anticipate the critics needs. Brutal as arbitrariness is, you can learn to defend your designs.</p><p>It also means that people start to understand what critics are saying underneath the base-level knee-jerk reactions. ‘That drawing is too abstract’ means they could not understand it enough to make an informed decision. ‘Where is the data to support this’ means you glossed over the rigorous research you’ve done. ‘Maybe you can look at this other person’s work’ means someone else communicated about their design effectively and you should not copy their work, but could try to emulate their presentation style.</p><h3>Roll</h3><p>Does this all sound super vague? Good.</p><p><strong>Roll</strong> <strong>with</strong> <strong>it</strong>.</p><p>Find yourself within the confines and constructs of your projects.</p><p>This is a real life conversation I once had with a client:</p><p>J: I need a website.</p><p>Me: awesome. Let’s start by looking at some sites you like. Tell me what you like about each of them.</p><p>J: this one is nice. I like it.</p><p>Me: cool. Can you tell me what you like about it?</p><p>J: it’s…nice?</p><p>This is not a joke. It’s real. And I did ultimately end up making a site for him. And he really liked it. But it took a lot of asking questions and iterating on features, looking at magazines and books and color palettes. Finding a common language where we could exist in the project. This is what design is. Understanding your client and user needs simultaneously and creating something that fits both. And if it sounds like getting clear answers is akin to pulling teeth, sometimes it is. And if that isn’t something you’re into, it can be a long bumpy ride ahead.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*742oIMknnoARoz_-KZPMpQ.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=172e7891c99b" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>