<?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 Josh Winters on Medium]]></title>
        <description><![CDATA[Stories by Josh Winters on Medium]]></description>
        <link>https://medium.com/@alwayswinters?source=rss-fb03954a566a------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*8F77s8_XOZH_K14EXz6Urg.jpeg</url>
            <title>Stories by Josh Winters on Medium</title>
            <link>https://medium.com/@alwayswinters?source=rss-fb03954a566a------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 00:52:16 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@alwayswinters/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[The New Pricing Model for Kapiche]]></title>
            <link>https://medium.com/kapiche/the-new-pricing-model-for-kapiche-3de00784b2cf?source=rss-fb03954a566a------2</link>
            <guid isPermaLink="false">https://medium.com/p/3de00784b2cf</guid>
            <category><![CDATA[product-updates]]></category>
            <dc:creator><![CDATA[Josh Winters]]></dc:creator>
            <pubDate>Wed, 11 Apr 2018 04:05:15 GMT</pubDate>
            <atom:updated>2018-04-11T04:06:18.883Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dFwBpMv8egOkgKFwIbp_-g.png" /></figure><p>Here’s what affects you, as an existing or future user of Kapiche:</p><ul><li><strong>Data Units are used for uploading data, rather than running analysis. </strong>We received feedback of customers hesitating to ‘try different analysis’ on their data because they didn’t want to use up Data Units. The aim of this change promotes experimentation, sharing, and exploration of data without restriction.</li><li><strong>One Data Unit = one row in your data source.</strong> When you upload a survey containing 10,000 responses, it will consume 10,000 Data Units.</li><li><strong>Your Data Unit balance has not changed. </strong>You’ll most likely get more value per Data Unit, so we thought it’d be nice to give existing customers a thank-you bonus rather than recalculating Data Units.</li><li><strong>Plans have changed.</strong> We now have three pricing tiers with monthly and yearly billing options (instead of just having one monthly and one yearly plan). We’ll be emailing existing users to request which plan they’d like to be switched over to. You can view the new pricing tiers <a href="https://kapiche.com/pricing">here</a>.</li></ul><blockquote>I’ve written a more detailed summary of the changes below, for those who are interested.</blockquote><h3>Data Units</h3><p>The old usage model was based on Text Excerpts, which were also called Data Units.</p><p><strong>The old Data Unit definition:</strong></p><blockquote>A Data Unit, or Text Excerpt, is a block of text up to 2 sentences long and doesn’t cross paragraph boundaries.</blockquote><p>There were a couple problems with this. Mainly, it’s really hard to predict and gauge your usage. Also, the definition is unecessarily complicated and was invented a long time ago based on an old product direction.</p><p><strong>The new Data Unit definition:</strong></p><blockquote>A Data Unit equates to one row in your data source.</blockquote><p>Upload an Excel table with <strong>1,000</strong> rows of data? That’ll use <strong>1,000</strong> Data Units.</p><p>Pulling in <strong>5,000</strong> responses from Qualtrics? That’ll use <strong>5,000</strong> Data Units.</p><p>Simpler, easier to manage, and more useful for the type of data our users tend to work with — free text survey responses.</p><h3>Data Unit Consumption on Data Upload</h3><p>We asked customers how they felt about using Data Units. The response was that it was mostly okay, however they were hesitant to try different analysis settings or let other departments run analysis for fear of using up Data Units.</p><p>We didn’t like that users were so scared to use up Data Units to run an analysis. We asked ourselves:</p><blockquote>What’s the best way to manage usage with as little impediment as possible?</blockquote><p><strong>Option 1: Restrict based on amount of data stored </strong>(similar to the Dropbox model). Perhaps a nice short term solution, but as a user’s data inevitably accumulates, it becomes unreasonable to pay to upgrade the storage amount. We also don’t want users to feel like they should delete potentially useful ‘old’ data to save on allocated space, especially considering we’re including more and more features based around tracking changes and trends in your data over time.</p><p><strong>Option 2: Restrict based on number of Projects. </strong>This turns out to be a very inconsistent model, as users may only ever need one Project with a ludicrous amount of data. Another user, however, may need many Projects with smaller amounts of data in each. There’s no way to tell which size company needs which number of Projects. Having no restriction at the data level wasn’t feasible from an economical standpoint, either.</p><p><strong>Option 3 (the winner): Restrict based on data upload. </strong>This was the option we chose. The principle is simple and easy to understand: you pay for the data you upload into Kapiche. This doesn’t have the problems the previous two options have; larger companies will generally have more data, along with the budget + return-on-investment for analysing that data. There’s no penalty for storing or accumulating a large dataset in Kapiche, and no drawback for running as many analyses you want on your data either. The downside is there’s a potential issue when users upload messy data, or accidentally upload the wrong data (but this problem existed in the old usage model, too). We’re trying to figure out a way to ensure our users upload clean data, but that’s a tricky problem to solve.</p><h3>Pricing Plans</h3><p><strong>The old pricing plans were only two plans (monthly and yearly) without much flexibility on usage.</strong></p><p>The new plans follow a more familiar structure: three tiers with monthly/yearly billing options. Not too much to write here — you can view the pricing plans <a href="https://kapiche.com/pricing">here</a>.</p><p>We hope these changes roll out smoothly; it’s no small thing to overhaul your pricing model and introduce new pricing plans. Even if it might be improvements all round, we expect some inertia to change. We are very, <em>very</em> open to talking to our customers. If you’re an existing customer with any concerns, don’t hesitate to <a href="https://kapiche.com/contact">contact us</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3de00784b2cf" width="1" height="1" alt=""><hr><p><a href="https://medium.com/kapiche/the-new-pricing-model-for-kapiche-3de00784b2cf">The New Pricing Model for Kapiche</a> was originally published in <a href="https://medium.com/kapiche">Kapiche</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Product Update — April 9th]]></title>
            <link>https://medium.com/kapiche/product-update-april-9th-caf43fbd117?source=rss-fb03954a566a------2</link>
            <guid isPermaLink="false">https://medium.com/p/caf43fbd117</guid>
            <category><![CDATA[net-promoter-score]]></category>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[nlp]]></category>
            <dc:creator><![CDATA[Josh Winters]]></dc:creator>
            <pubDate>Mon, 09 Apr 2018 00:38:50 GMT</pubDate>
            <atom:updated>2018-04-09T00:38:50.524Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dFwBpMv8egOkgKFwIbp_-g.png" /></figure><p><em>This update brings two new highly requested features and a big performance increase in sentiment processing speed.</em></p><h4>Mixed Sentiment</h4><p>The world of sentiment analysis can be a <a href="https://blog.kapiche.com/four-reasons-sentiment-analysis-is-misinterpreted-4d9bb59b41b9">little confusing</a> sometimes, but one thing is for certain, only having 3 labels for sentiment (positive, negative, and neutral) isn’t enough. What about the instances where a customer is being positive and negative in the same response?</p><p>Included as part of this release is <strong>mixed sentiment</strong>, designed to address this problem. Now when a customer response is both positive and negative, it will be marked as <strong>mixed</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/238/1*09vw1oB-C7ytiWgQ6tJYXg.png" /><figcaption>New tooltips showing sentiment values.</figcaption></figure><p>We’ve updated all pages where Sentiment is shown as well as an expanded tooltip on the Storyboard that gives you a full sentiment breakdown.</p><p><strong><em>Important: Mixed sentiment will only affect new data uploaded to projects, not existing data!</em></strong></p><h4>Timeline Trend</h4><p>Before this update, you could only view the trend of topics on a separate timeline screen. Now each Topic/Segment/Term screen has its own timeline trend component.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mDWxCsPAFFghNWK00RwhAg.png" /><figcaption>The new Timeline component showing the trend for a Topic</figcaption></figure><p>This new feature allows you to track the trend over time of any Topic, Segment or Term alongside the existing information for each of these screens. It will only be visible when you have date information, and will update as you add new data to your Project.</p><h4>Performance Increase in Sentiment Processing Speed</h4><p>Assigning sentiment to data is the slowest part of the data upload process. This was a very annoying issue for our users (and us), so we’ve made it faster… much faster. Our tests show a <strong>5-10x speed increase</strong> in data processing time.</p><h4>Bug Fixes</h4><ul><li>Fixed some minor navigation issues around Qualtrics integration</li><li>Project errors occur on the current screen rather than at the end of the Create Project process</li><li>Fixed an NPS component loading error.</li></ul><h4>Final Notes</h4><p>Big thanks to the input from our customers. Your feedback is what drives Kapiche forward! Don’t ever hesitate to <a href="https://kapiche.com/contact">contact us</a> or email me directly at <a href="/service/mailto:josh.winters@kapiche.com">josh.winters@kapiche.com</a> if you have questions, suggestions, or would just like to learn more about Kapiche.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=caf43fbd117" width="1" height="1" alt=""><hr><p><a href="https://medium.com/kapiche/product-update-april-9th-caf43fbd117">Product Update — April 9th</a> was originally published in <a href="https://medium.com/kapiche">Kapiche</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[NPS Surveys: Stop asking “What” and ask “Why”]]></title>
            <link>https://medium.com/@alwayswinters/nps-surveys-stop-asking-what-and-ask-why-29e44ad46633?source=rss-fb03954a566a------2</link>
            <guid isPermaLink="false">https://medium.com/p/29e44ad46633</guid>
            <category><![CDATA[text-analytics]]></category>
            <category><![CDATA[net-promoter-score]]></category>
            <dc:creator><![CDATA[Josh Winters]]></dc:creator>
            <pubDate>Tue, 27 Mar 2018 01:24:27 GMT</pubDate>
            <atom:updated>2018-03-27T06:29:59.826Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2-q0JoE87F6z2Xdg8Jjcdg.png" /></figure><p>Ever since Frederick F. Reichheld (the creator of NPS) published his article on NPS in <em>“The One Number You Need to Grow”</em> in the Harvard Business Review, NPS has become a growing trend in customer loyalty and customer relationship management. It’s continuing to gain traction, and is still a hot topic of discussion (and debate) around the business community.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/552/1*YrAykldf-2U6LCkQ9yx-Gw.png" /><figcaption>A typical NPS survey.</figcaption></figure><h4>A quick recap on NPS</h4><p><em>Skip this section if you’re already familiar with NPS.</em></p><p>Most NPS surveys use a scale that range from 0 to 10, along with an open ended question asking the reason for the score.</p><p>Clients who respond with a score of 9 to 10 are called Promoters. These are loyal customers who repeatedly buy from your company and recommend others to do the same. They’re more likely to forgive mistakes with your service or product, and are more engaged with your company brand.</p><p>Those that respond with 7 and 8 are called Passives. Passives<strong> </strong>are satisfied but unenthusiastic customers who can be lured by the competition to move or try other products and services. If they find new products/services interesting they will move without much hesitation.</p><p>Detractors are unhappy customers who are ‘at risk’ because of their perception about your company, and are those who spread the majority of negative word of mouth. They will give scores ranging from 0 to 6.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*efl4bIEJbkKRnh84c5CxtA.png" /></figure><p>Net Promoter Score is calculated by subtracting the percentage of Detractors from the percentage of Promoters. It gives a good metric to track and work towards improving.</p><p>Simple enough, right? So where do some companies trip up?</p><h4>The value is in the open-ended question</h4><p>The open-ended question is where all the goodness lies — not the number which so many companies blindly chase after. While a bit over the top, <a href="https://medium.com/ux-immersion-interactions/get-a-better-ux-metric-from-your-nps-survey-data-622ed0ad2ce3">Jared Spool says it quite humourously</a>: <em>“That number the survey respondents gave has no value. It’s like the skin of the mango. There’s nothing good about it. Just throw it away.”</em></p><h4>Asking “Why” not “What”</h4><p>The simple act of asking <strong><em>why </em></strong>the customer has given you their score is paramount in making NPS more than just a number and something to actually take action upon. It’s a small wording difference that triggers different responses from people. You’re much more likely to get useful, descriptive feedback from someone when you ask why they gave their score instead of what the reason was for their score. Asking “Why” encourages people to elaborate on their reasons rather than just stating what it is. Here’s a simplified example to illustrate the point:</p><p>Asking <em>“what”</em> can result in a rather unsubstantial response because it narrows the scope of the question:</p><blockquote>“What is the reason for your score?” <strong><em>The fast service.</em></strong></blockquote><p>Asking <em>“why”</em> directs the person to elaborate more on their reason and broadens the scope:</p><blockquote>“Why did you give us that score?” <strong><em>Because the service is faster than other stores.</em></strong></blockquote><p>Getting richer and more descriptive qualitative data is extremely important; better data collection means better results out of your <a href="https://blog.kapiche.com/why-understanding-voice-of-the-customer-is-critical-for-your-business-success-59c27c0749e3">voice of customer analysis</a>. And yes, processing large amounts of long, free-text data can be a hard and time-consuming task. <strong>Well, it <em>used </em>to be.</strong></p><h4>Technology moves fast</h4><p>In a world where we didn’t have the analytics technology we have now, asking <em>“what”</em> probably made life easier; it makes manual coding of surveys faster because the feedback is more likely to be short, quick statements which are easy to bucket into categories. Historically, analysing text data has been notoriously difficult, but today we have access to a whole range of techniques, algorithms, and technologies that make it automated, more accurate, less biased, and extremely fast. We can afford to aim for long, descriptive answers to customer feedback nowadays. Not just aim for, but actively seek it. The payoff is huge, and companies live or die by understanding their customers.</p><p>Our next post will be on text analytics methods, which will go further into explaining the technologies behind them. Hope to see you then!</p><p><em>Kapiche was built for companies who want to empathize with their customers at scale. Our multi-structured analytics software helps companies truly understand what their customers are trying to tell them. Learn more at </em><a href="http://pxlme.me/The4bpBH"><em>kapiche.com</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=29e44ad46633" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Going Beyond Word Clouds: Why We Built The Storyboard]]></title>
            <link>https://medium.com/@alwayswinters/going-beyond-word-clouds-why-we-built-the-storyboard-101a8c61519c?source=rss-fb03954a566a------2</link>
            <guid isPermaLink="false">https://medium.com/p/101a8c61519c</guid>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[product-design]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[text-analytics]]></category>
            <dc:creator><![CDATA[Josh Winters]]></dc:creator>
            <pubDate>Mon, 19 Mar 2018 22:01:02 GMT</pubDate>
            <atom:updated>2018-03-20T00:58:37.605Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3L4cXEXvhoZZFudE82FDhw.png" /></figure><p>Kapiche had a question: what’s the best way to help a new user begin to understand the text data they upload? How could we transform a large, complex dataset into a compelling visualization that presents a narrative and conveys the scope of the data?</p><p>Did someone say… <strong>word cloud?</strong></p><p>A word cloud fits the description at first glance; it’s a nifty little visualisation that shows the keywords of a textual dataset with words sized based on frequency. You’ve probably seen them before — on blogs, news articles, and the left side of this article’s header image.</p><h3>Word clouds only go so far</h3><p>As <a href="http://www.niemanlab.org/2011/10/word-clouds-considered-harmful/">Jacob Harris writes</a>, word clouds throw out all of the principles good data journalism strive for:</p><blockquote><em>A narrative that pares away extraneous information to find a story in the data; context to help the reader understand the basics of the subject; interviewing the data to find its flaws and be sure of our conclusions.</em></blockquote><p>Word clouds probably never sought out to do these things, mind you. They are simply a visually appealing and accessible way to display text frequency. How people misuse them is not what we’re here to talk about, that’s already been <a href="https://towardsdatascience.com/can-we-please-stop-using-word-clouds-eca2bbda7b9d">discussed enough</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/576/1*nydqKQCSagGhDsIU4Xk2Jg.png" /><figcaption>I spent more time than I’m proud of to make it look this good. Made with <a href="https://worditout.com/">Worditout</a>.</figcaption></figure><p>Word clouds are the <em>opposite </em>of what we want to achieve at Kapiche; we want our users to be <a href="https://blog.kapiche.com/four-reasons-sentiment-analysis-is-misinterpreted-4d9bb59b41b9">accurate</a>, explorative, and critical when analyzing their data. We want to give them the tools to understand the whole narrative and the context surrounding what people are talking about. So if wordclouds aren’t the solution, then what <em>is</em>?</p><h3>Enter the Storyboard</h3><p>We knew word clouds weren’t enough, so we set out to build something better. In particular, we aimed to:</p><ul><li>Provide a starting point — a big picture view — for users to orient themselves and begin delving into a dataset;</li><li>Provide enough context and useful information without overwhelming the viewer;</li><li>Invite interaction and exploration to promote understanding of a dataset; and</li><li>Avoid injecting bias via aesthetic choices or design elements.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Dpkev8JjrFRr8ztcpJDjJQ.png" /><figcaption>The Storyboard takes the visually exciting, accessible nature of word clouds and the depth &amp; structure from <a href="https://en.wikipedia.org/wiki/Topic_model">topic models</a> to display large amounts of complex text data in an easily digestible and informative way.</figcaption></figure><h4><strong>Breakdown of the Storyboard</strong></h4><p>Glancing at the above image should give you a general understanding of what is going on in this anonymous dataset, even without being told anything. It looks like a concept map, but we’ve added some interesting design choices to add understanding and usefulness:</p><p><strong>Let’s break it down:</strong></p><ul><li><strong>Terms form topics</strong>. For example, <em>“friendly, helpful, staff”</em> is a topic. An example of a term would be <em>“staff”. </em>All of the terms are actual words in the dataset — we do not introduce any human bias by categorizing through codebooks or dictionaries (which also means there’s no tedious setup required, either). It’s all done automatically through unsupervised machine learning algorithms.</li><li><strong>Topics &amp; terms are colored &amp; sized by frequency</strong>, from purple (highest) to yellow (lowest). The size of the circles behind the terms also helps to reflect frequency. You can see in the above image that <em>“friendly, helpful, staff”</em> and <em>“good, quality, products, brands, high”</em> are the highest frequency topics.</li><li><strong>Topics &amp; terms are positioned by relatedness</strong>. For example, in the above dataset we can see that <em>“good, quality, products, brands, high”</em> is related to <em>“price”</em>. On the contrary, <em>“price”</em> is not likely to be related to <em>“customer, service”.</em></li></ul><p>By using the design principles of colour, size, and position to distinguish different meanings in the data, our users navigate the storyboard very quickly and then use it as a launch point into deeper insights.</p><p>Clicking into a topic sends you to a different screen which shows the full details of that topic, including sentiment, co-occurences with other topics, topic drivers, NPS data, data coverage stats, and more. This screen also shows you the expanded topic with all of the terms included, which we call the Context Network, shown here:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/483/1*ucJGq0wDvucTkMONDnysdQ.png" /><figcaption>Context Network showing the full composition of the <em>“friendly, helpful, staff” topic.</em></figcaption></figure><p><strong>One major advantage is that all of this happens within a few minutes of uploading a dataset.</strong> You don’t have to wait days or weeks for someone to process the data, because a lot of people at Kapiche put countless hours into figuring out how to let you generate it from your data automatically. This also means that the size of the data source doesn’t cause it to take longer to generate — putting 50,000 survey responses through Kapiche won’t take much longer than 1,000.</p><h3>Final thoughts</h3><p>We’ve received a lot of good feedback on the Storyboard, and it has been achieving the goals we set for it. It’s an excellent starting point to begin understanding the narrative of a dataset, and the interactivity makes it a powerful springboard for delving deeper into interesting points of the data. There are no random aesthetics — every design element has a purpose and conveys something meaningful. I’m sure there are more improvements to come (<em>everything </em>can be improved) as we see how users interact with the product as a whole.</p><p><em>Kapiche was built for companies who want to empathize with their customers at scale. Our multi-structured analytics software helps companies truly understand what their customers are trying to tell them. Learn more at </em><a href="http://pxlme.me/The4bpBH"><em>kapiche.com</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=101a8c61519c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Product Update — February 2nd]]></title>
            <link>https://medium.com/kapiche/product-update-february-2nd-18e122b89faf?source=rss-fb03954a566a------2</link>
            <guid isPermaLink="false">https://medium.com/p/18e122b89faf</guid>
            <category><![CDATA[customer-experience]]></category>
            <category><![CDATA[net-promoter-score]]></category>
            <category><![CDATA[text-analytics]]></category>
            <category><![CDATA[voice-of-the-customer]]></category>
            <dc:creator><![CDATA[Josh Winters]]></dc:creator>
            <pubDate>Fri, 02 Feb 2018 04:02:25 GMT</pubDate>
            <atom:updated>2018-02-02T04:02:25.200Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*v-1TfhVx6Fp3ggpv9XGIfQ.png" /></figure><p><em>This update brings multi-segment filtering, a handful of improvements, and the beginnings of some exciting new features around NPS-specific data.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vOr7vbekg6Rk_9am8Pp66Q.png" /></figure><h4>View &amp; Compare Segments</h4><ul><li>Multiple Segments can now be used in a single query!</li><li>Any time you select two segments of a single variable, it will result in an OR of those segments. In other words, if you search for <strong>{Gender: Male}{Gender: Female}</strong> it will show results which had either of those values (i.e. an OR function). This is different to searching <strong>{Gender: Male}{NPS: 10}</strong>, which would only show results that include both of those segments <strong><em>together </em></strong>(i.e. an AND function).</li><li><strong>Compare Segments now has a combined Topic Comparison chart with correlation/frequency view toggle.</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*H-KoBkn0CIlhB-g7xyzlAQ.png" /><figcaption><strong><em>Comparing differences between your Segment queries is easier now.</em></strong></figcaption></figure><ul><li>Segment screens now have an improved Segment Comparison chart, which is now filtered by a single Segment. You can still see the old “Top Overall” segment comparison in the filter dropdown.</li><li><strong>Compare Segments now has a fixed query bar, and will always be visible while scrolling.</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*McE2QelCa3ZAm67PWV-X5w.gif" /><figcaption><strong>You can now always see the queries you’re comparing on the Compare Segments screen.</strong></figcaption></figure><h4>Projects</h4><ul><li>Analysis status messages while it’s being processed in the Project Overview screen are more meaningful.</li><li>To create a new Project, you now need at least one text field selected and enough text data to form a useful analysis model.</li><li><strong>There is now an NPS field type:</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/238/1*lEF1m5iDBrG2ENBC5P-zlQ.png" /><figcaption><strong>NPS field type on the Schema screen.</strong></figcaption></figure><ul><li>NPS fields will be used in a later update to automatically create Promoter, Detractor, and Passive Segments for easier queries (and other cool things!). In the meantime, we automatically calculate the values and show them in the Analysis Settings. This NPS panel will be making its grand appearance on the new Summary screen and View Segment screen in the future!</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/850/1*oOBVTi4nkUe9WCKocNh9LA.png" /><figcaption><strong>If you upload data with the NPS field type, you’ll get the NPS Summary in the Analysis Details screen.</strong></figcaption></figure><h4>Data Visualisations &amp; Charts</h4><ul><li>Gridlines are slightly darker to improve visibility.</li></ul><h4>Analysis Settings</h4><ul><li>French stopword list is now available.</li></ul><h4>Final Notes</h4><p>Thank you to all our testers for the feedback &amp; time given. There are still many more features and improvements we’re working on, so stay tuned!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=18e122b89faf" width="1" height="1" alt=""><hr><p><a href="https://medium.com/kapiche/product-update-february-2nd-18e122b89faf">Product Update — February 2nd</a> was originally published in <a href="https://medium.com/kapiche">Kapiche</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Product Update — December 21st]]></title>
            <link>https://medium.com/kapiche/product-update-december-21th-a26f1cc8c7f0?source=rss-fb03954a566a------2</link>
            <guid isPermaLink="false">https://medium.com/p/a26f1cc8c7f0</guid>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[product-updates]]></category>
            <category><![CDATA[nps]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[voice-of-the-customer]]></category>
            <dc:creator><![CDATA[Josh Winters]]></dc:creator>
            <pubDate>Thu, 21 Dec 2017 06:28:11 GMT</pubDate>
            <atom:updated>2018-01-11T23:17:40.701Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*v-1TfhVx6Fp3ggpv9XGIfQ.png" /></figure><p><em>December brings one of the biggest product updates to Kapiche so far. We’ve updated the Storyboard, introduced a better Project workflow and added a multitude of small quality of life changes.</em></p><p><strong><em>Update:</em></strong> The December product update webinar is available on <a href="https://www.youtube.com/watch?v=ccfNGXg8Kzw">Youtube</a>!</p><h4>Projects, Data Collections, and Analyses</h4><p>The most critical part of this update is the new workflow in which you manage your Projects and the data within them. The concept of Data Collections are gone, and is now just “Data” within a Project. New analyses use all of the data in the Project, and you can update existing analyses with new data.</p><ul><li><strong>Existing Data Collections will be converted into Projects (renamed according to the old Project and Data Collection name). This process is automatic.</strong></li><li><strong>You will need to update each existing Analysis.</strong></li><li><strong>Topics in existing Analyses may change once updated.</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dSpZCGUwgW9kBPHvxrVzOw.png" /><figcaption>Data and Analysis are linked closely within each Project.</figcaption></figure><ul><li>You now upload a data file when creating a new Project. The first data file you upload dictates the “Schema” of your Project. This is the table/column structure all the files in your Project must follow if you add new data.</li><li>New analyses you create use <strong><em>all of the data</em> </strong>in your Project. With this new workflow, think of Projects as the old “Data Collections”. You’ll likely be creating more Projects now, and have more control over what happens in each Project.</li><li>Any time you add new data, you can update any existing analyses to use that data. Updating an analysis means the topics &amp; storyboard will <strong><em>not</em></strong><em> </em>be modified (and any new topics will be disregarded), so you can see how the frequencies of topics have changed with new data. You can of course create a new analysis which will generate a topic model with the new data as well.</li><li>It is now more obvious when data and analyses are being processed, as well as what you can and can’t do whilst they are processing.</li></ul><p>We’ll be user testing and monitoring this feature closely and iterating on the feedback we receive. <strong>Contact us if you need help at </strong><a href="/service/mailto:hello@kapiche.com"><strong>hello@kapiche.com</strong></a></p><h4>General</h4><ul><li>Topics have a new colour scheme based on frequency, from Yellow (highest) to Purple (lowest).</li></ul><h4>Storyboard</h4><ul><li>Position of topics is fixed (they stay in the same position every time you load the visual), and the Storyboard now loads with less erratic movement.</li><li>There is a new topic frequency chart on the right side of the screen.</li><li>You can now edit the granularity of topics through the “Edit Storyboard” button. This doesn’t just impact the visuals — the rest of your analysis will use the new topics, too!</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/895/1*ztzixkcH1k4H-t2o-_KaWg.png" /><figcaption>The Storyboard now has a slider (Edit Topics button) that lets you adjust Topic granularity.</figcaption></figure><h4>Data Visualisations &amp; Charts</h4><ul><li>You can now find a .csv export option for any chart that uses table data.</li><li>Bar charts now have a minimum bar height.</li><li>Topic Correlations on the Topic page is now a Relative Frequency chart.</li><li>The Context Network when viewing a topic now has reduced text overlap and better interaction.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/557/1*sKuuh_7FzrTJPVxD-LSEVw.png" /></figure><h4>Analysis Summary</h4><p>We have removed the Summary screen for now. You’ll now be greeted with the Storyboard when you open an Analysis. We may re-introduce the Summary screen at a later time when it is at a place we’re happy with.</p><h4>Segments</h4><ul><li>Sentiment is now created for each analysis as a Variable, with positive/neutral/negative segments.</li><li>A column can have up to 25 unique values to become a Variable (up from 15).</li><li>Multi-segment queries are coming in the next update.</li></ul><h4>Final Notes</h4><p>Thank you to all our testers for the feedback &amp; time given. There are still many more features and improvements we’re working on, so stay tuned!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a26f1cc8c7f0" width="1" height="1" alt=""><hr><p><a href="https://medium.com/kapiche/product-update-december-21th-a26f1cc8c7f0">Product Update — December 21st</a> was originally published in <a href="https://medium.com/kapiche">Kapiche</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>