<?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 Abi Gurumurthy on Medium]]></title>
        <description><![CDATA[Stories by Abi Gurumurthy on Medium]]></description>
        <link>https://medium.com/@abi_g?source=rss-61f3fcd55fe4------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*WFt9ABJj9xCNSLsgtKAs-w.png</url>
            <title>Stories by Abi Gurumurthy on Medium</title>
            <link>https://medium.com/@abi_g?source=rss-61f3fcd55fe4------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 05:44:13 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@abi_g/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[Hypothesis Testing]]></title>
            <link>https://medium.com/@abi_g/hypothesis-testing-137208c6421c?source=rss-61f3fcd55fe4------2</link>
            <guid isPermaLink="false">https://medium.com/p/137208c6421c</guid>
            <category><![CDATA[design-thinking]]></category>
            <category><![CDATA[hypothesis-testing]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[user-research]]></category>
            <dc:creator><![CDATA[Abi Gurumurthy]]></dc:creator>
            <pubDate>Wed, 07 Aug 2019 03:38:58 GMT</pubDate>
            <atom:updated>2019-08-07T03:38:58.108Z</atom:updated>
            <content:encoded><![CDATA[<p>Every product manager would have come to the crossroads in deciding if something is a problem worth solving for the customers.</p><p>Although there are many channels to verify a problem statement, hypothesis testing has been very useful to accomplish the task and move forward with building the feature/product.</p><p>I recently came across a case study that I thought would be worth sharing to the wider audience. I was trying to gather customer interest in building a feature for the <a href="https://transitapp.com/">Transit App</a> where the user can purchase tickets within the app and not rely on external kiosks, online portals or stand-alone fare machines.</p><p>The idea is to spend the least amount of time possible but also gather some data points to have some indication if this is a problem worth solving. I am sharing the way I thought would be better to approach this.</p><p><strong>Summary</strong>: This project aims to run a hypothesis-driven validation of users interest in purchasing public transit tickets (bus fares) within the app.</p><p>The goal is to gather enough data and measure metrics to validate the idea before wasting time and money building a solution to a problem that might not exist. We should expect more than 60% users providing a positive response and prove our hypothesis right for this experiment to be successful.</p><p>Note: 60% is subjective and it depends on the sample space. The assumption is that the app is being used by more than a million users today.</p><p><strong>Test Goals</strong></p><p>1. Gauge interest from users if they would like to purchase ticket(s) within the app.</p><p>2. Measure activities that are related to purchasing tickets within the app.</p><p><strong>Problem</strong></p><p>No easy way to purchase transit tickets: Commuters using the transit app have to go through multiple hoops to purchase a ticket regardless of the mode of transportation. There is no single way to plan, track and pay for trips within the same ecosystem.</p><p>There is not a single app that could interact with individual services (trains, buses or subway) to calculate the fares associated with each segment and present a total, integrated fare for the trip, let alone purchase it through a single app.</p><p><strong>How is it solved today?</strong></p><p>Most common ways to purchase a ticket currently are either through a kiosk at a station, fare machines, multiple online portals, or through a vendor.</p><p><strong>Who is my customer?</strong> We have two major personas.</p><p>1. As a local, I want to get from point A to B and also purchase a ticket through the app.</p><p>2. As a tourist, I would like to explore specific points of interest and have all my mobility options accessible in one place (including purchasing tickets).</p><p><strong>Assumptions</strong></p><p>1. All users using the transit app are considered equal. We are not splitting active versus passive users.</p><p>2. The initial rollout of this test will be in the top 5 cities in the US (namely NYC, LA, SF, Austin etc.) where there are widespread public transit systems.</p><p>3. Survey emails can capture the interest in users without any development effort, but we do not want our users to go through the pain of taking a survey. Even enterprise customers shun away from filling a survey unless necessary.</p><p><strong>Edge cases / Caveats</strong></p><p>1. An overwhelmingly positive response to this test doesn’t necessarily mean users will be using the feature in the future. Filling in credit card details within an app is something that most users don’t like doing compared to swiping a credit card in a kiosk.</p><p>2. This experiment only validates our idea, we will still need to run usability studies once our hypothesis is confirmed. E.g.: What are the preferred method of buying a ticket?- PayPal, G pay, Apple pay Amazon pay etc.</p><p>3. Transit app does not currently have a walled garden. Mobile transit ticketing will require users to sign in thereby driving them away. The churn of users is something we need to worry about.</p><p>4. Depending on the city, we need to support multiple modes, operators, and ticketing providers.</p><p>5. A user with no digital services will not be able to purchase tickets through the app. We are not solving the pain point of a cash transaction.</p><p>6. A user purchasing an incorrect pass cannot upgrade to a different one.</p><p>7. Different ticketing agencies use different vendors. Onboarding them requires effort.</p><p><strong>Proposal</strong></p><p>The idea is to be non-intrusive to the existing/new users and gather data from them.</p><p><strong>User Experience / Journey</strong></p><p>1. A user opens the transit app and the lower bottom of the screen has a question in the form of evaluating the user’s interest. The user can take one of three actions — thumbs up, thumbs down or close the section. Shown in the wireframe below (Figure 1 and 2)</p><p>2. This question only appears once for a particular account or a guest user. After an action is taken, it will never appear again.</p><p><strong>Design Scenarios</strong></p><p>1) When the user taps the “thumbs up” or “thumbs down” option, display an auto disappearing message “Thank you for your feedback!”</p><p>2) When the user taps the “close” option, do not display any message.</p><p>3) A Thumbs up is considered a positive response.</p><p>4) A Thumbs down is considered a negative response.</p><p>5) Note to developer: Mixpanel can be used to track the clicks with relative ease and little to no dev effort.</p><p><strong>Metrics to measure</strong></p><p>1. A total number of positive and negative responses during a 30-day period.</p><p>2. Percentage of positive and negative responses with respect to DAU / WAU / MAU (Daily, Weekly, Monthly active users)</p><p>3. A total number of users clicking the “Fares” option under “Settings” after the start of the experiment. A user checking Fares would most likely seek to purchase a ticket in an outlet / online next.</p><p><strong>Wireframes</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/891/1*mAZCudHUs2I4rAprvjOhkQ.png" /><figcaption>The survey question</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/917/1*I4b2_-3DI2mRhEPylMNcLg.png" /><figcaption>Survey submission UI feedback</figcaption></figure><p>Please share ideas, thoughts and your experience with hypothesis testing on consumer products.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=137208c6421c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[LET ME GIVE YOU A ‘SCOOP’]]></title>
            <link>https://medium.com/@abi_g/let-me-give-you-a-scoop-b5633e0876dd?source=rss-61f3fcd55fe4------2</link>
            <guid isPermaLink="false">https://medium.com/p/b5633e0876dd</guid>
            <category><![CDATA[silicon-valley]]></category>
            <category><![CDATA[uber]]></category>
            <category><![CDATA[product]]></category>
            <category><![CDATA[san-francisco]]></category>
            <category><![CDATA[scoop]]></category>
            <dc:creator><![CDATA[Abi Gurumurthy]]></dc:creator>
            <pubDate>Thu, 04 Oct 2018 04:45:33 GMT</pubDate>
            <atom:updated>2018-10-04T04:59:03.460Z</atom:updated>
            <content:encoded><![CDATA[<p>As a silicon valley resident, I am one among the thousands of people losing parts of their valuable time sitting on the clean Asphalt<strong> </strong>roads and starting at the never ending queue of vehicles in front of me only to wonder how <a href="https://en.wikipedia.org/wiki/Henry_Ford">Mr.Ford</a> would think in today’s world about creating faster lanes and not just faster cars.</p><p>I recently signed up and started using <a href="https://www.takescoop.com/">Scoop</a>, which is an app used by daily commuters to share their rides and car pool depending upon their distance, time, location and immigration status (obviously kidding). You can either ride or drive based on your availability and needs.</p><p>Scoop matches you with a fellow driver/rider who is planning to start/end their day around the same time window as you. You then get notified about the location / phone number of the person and you can sync up through the app or directly to notify the person.</p><p><strong>What is unique? </strong>Car pooling, Improves user commute, engagement between driver and rider by meeting new local people and the overall experience of having a conversational ride (for people commuting alone).</p><p>TADA. A genuine solution to the commuter woes.</p><p>As much as I like the product and its necessity, there are multiple ways it needs to be iterated to improve user needs.</p><p><strong>Finding the best match</strong></p><p>Scoop always lets you have a start and end destination along with a time window to match other users wanting to ride or drive with you. Once the deadline for creating a request is complete, you cannot change the From/To addresses but the time window can be adjusted. Although it makes sense to lock in the addresses, for users who are not being matched, this is a painful solution since I as a driver, if not matched with a rider, would like to change the ‘From’ address so that I can pick up my partner on the way and still look for rides in the same direction (To address) from the mid point.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/840/1*mSOx0TXVFJTPnbbktA6SkQ.png" /></figure><p>This way, I can still pick up a rider from the mid way (say, San Carlos) and the time window for a new user to ride also becomes a little flexible. I completely understand that Scoop doesn’t want to be Uber but this functionality would definitely improve usage and better the experience for users.</p><p><strong>Real Time Navigation</strong></p><p>One of the greatest pain points is to not be able to see if the driver has started from their location or is in transit. The app shows a map of the location with a car that stands still at the starting point of the driver. This gives a lot of anxiety when you schedule scoop as a last mile or an intermediate transit coupled with a public transportation. Regardless of that, every user needs to know the current status of their pickup to have a good user experience.That brings me to my next point about the app’s speed of response. Currently, it’s incredibly slow in completing simple tasks that are mandatory to schedule a ride. I wondered if this was a device issue and found that it was platform independent.</p><p><strong>Phone numbers</strong></p><p>This is not something new. Masking the phone numbers and making it more private would be a better solution once Scoop scales. Privacy is an issue for any user centric product and scoop is no exception. I would not want my number to be available with random strangers. Of course it we end up having a great conversation and are meeting outside of work for leisure, that’s a different story.</p><p><strong>Tik Tok</strong></p><p>I cannot stress the importance of being on time. Scoop, for its part, does its best to give the driver a time to start so that they can reach the rider at the right pick up time and be on their way without any delays (I am not sure if it accounts for the traffic conditions at peak hours).</p><p>On the flip side, rider also gets a time where they need to wait outside to be picked up.</p><p>Scoop, however, underestimated the human behavior which has evolved over centuries. We, humans, tend to be late when there is no accountability. I missed my BART twice because the driver was late for reasons unexplained. Scoop unlike Uber does not have an option to fine/penalize the user for this behavior which extrinsically would encourage users to be on time.</p><p>The same holds good for the rider as well. I am unsure of an optimal solution to this problem but adopting a price / fellow user feedback based program would be a good start.</p><p><strong>+ 1</strong></p><p>Providing an option for letting users choose one or more riders would definitely be helpful in highly dense environments (specifically SF / San Jose). Most users drive a sedan (3+1) and giving them the option of picking up more than one rider in the same direction (without adding much distance) would be a win-win-win situation.</p><p><strong>The Mundane stuff</strong></p><p>Some people have a standard routine and it would be a great option to let the users fix their time window for a set period of time (say, a month) and not notify them to make a request every day/week. This makes their life much more easy (although mundane :P) provided they have a standard schedule.</p><p>All said and done, I, for a fact, know that this product alleviates a lot of end user commuter problems in today’s world and even small iterations would only provide a richer experience to the user. I can see myself coming up with more issues that I will come across as a customer and would love to hear from you from the UX stand point.</p><p>What would make Scoop click for you? Share your experiences on using the product.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b5633e0876dd" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>