Duetto Women Win Hackathon!

Over the last 10 days, 4 amazing ladies and I formed a team and participated in the #womenhackthecrisis - A Global Virtual All-Women Hackathon. The goal was to provide solutions to COVID-19 related challenges faced by essential workers, small businesses, and/or women seeking shelter. We pitched our idea last night to an incredible group of judges and an audience of other amazing teams. We are beyond excited to share that our team, named Hotel Haven, won first place!!!

Hotel Haven is a resource that aims to connect individuals, in need of a temporary home, with hotels in their area. In addition, donors can "gift" rooms to those in need and offer them comfort during challenging financial times.

The team worked incredibly hard outside of work and I couldn't be more proud! I'd like to personally give a special mention to each of my team members Jessica ClaytonEmily JesseLaura Slepetis, and Uma Parkala for their dedication and commitment to this challenge. You ladies are what it is all about!!

Thank you to Duetto for sponsoring our team and to all of our #duettocustomers for inspiring our idea. It truly brought us all together. Lastly, I'd also like to give a special mention to Nathan Gray from Coast Hotels for providing us with great information during our discovery phase of development! His team is doing amazing things with their Coast Cares rate and it sparked a lot of ideas for us. We hope we can share our idea with more hoteliers and start working on the next steps of Hotel Haven's future. Let's see where this can truly go!

Source: https://www.linkedin.com/pulse/duetto-wome...

Choosing the Right User Event Tracking Software

Before Getting Started…

I recommend that you read my article on Product Usability and Engagement Metrics. The most important thing to do before you begin evaluating different vendors is to make sure you have clearly identified the questions and hypotheses you would like answered. Every software company is different, with their own use cases and user needs.

At two different companies, I have rolled out user event tracking software. I’ve gone through a lot of trial and error, but ultimately these tools are the best weapon a Product Manager can have. Remember, we are NOT the user. Having an unbiased measurement of how our users are performing is truly the only way we can measure our successes and failures.

Tool Summaries

Note that the summaries below were researched and put together in early 2018. I am not affiliated with any of these companies and am only providing my personal opinion. My few bullets on each do not do each company justice, but my hope is that it at least gives you a guide for what solutions are out there and the subtle differences between them.

  • Mixpanel - Event Tracking + Analytics 

    • Requires events defined up front and engineering resources to add each new event moving forward

    • Events are not retroactively captured (UPDATE: This appears to have changed in 2019 with their latest release but I have not personally tested this)

    • Optimized for mobile app A/B testing

    • Clean interface for advanced analytics

    • HQ in San Francisco, CA

  • Amplitude - Event Tracking + Analytics 

    • Requires event tracking defined up front and engineering resources to add each new event moving forward 

    • Events are not retroactively captured

    • Caters primarily to an in-depth analysis of user behavior

    • Strong in identifying users cross-device

    • HQ in San Francisco, CA

  • Segment - Middleware

    • Consider Segment if you want to plug and play a number of tools across multiple departments

    • Built for developers

    • Cross-channel analytics 

    • HQ in San Francisco, CA

  • FullStory - Rewind customer sessions + Analytics

    • Watch full video playbacks of your user sessions

    • It masks PII data (personal information or anything you dub as sensitive data)

    • Very powerful for understanding user flows

    • HQ in Atlanta, GA

  • Heap - Tracking + Analytics 

    • Does not require events defined up front - Client side event tracking handled by them

    • Allows for retroactive event tracking from deployment date

    • Features:

      • Tagging (create individual events)

      • Grouping (organize events)

      • User segmentation

      • Funnels

    • HQ in San Francisco, CA

  • Pendo - Tracking + Analytics + Guides

    • Does not require events defined up front - Client side event tracking handled by them

    • Allows for retroactive event tracking from deployment date

    • Analytics Features:

      • Tagging (create individual events)

      • Grouping (organize events)

      • User segmentation

      • Funnels

      • Paths

    • Guide Features: 

      • Poll Responses

      • Guides - Customizable tooltips, Guide House (similar to Intercom bubble that lives on the page with unique content), survey options (ex: in app NPS), and Walkthroughs (interstitials)

    • HQ in Raleigh, NC

  • WalkMe - Guides + Tracking

    • Provides very sophisticated in-app user guides and tutorials

      • Note: It can become quite complex to maintain overtime

    • Ability to add tool tips, shout-outs, step-by-step guides and more

    • Provides tracking metrics on guides usage

    • Weak on user event tracking - they are moving into this realm, but it is not their bread and butter

    • Allows you to play back user sessions

    • HQ in San Francisco, CA

Ultimately, I chose Pendo for my user event tracking needs. I wanted a tool that required very little reliance on my engineering team to implement and maintain, so that eliminated Mixpanel and Amplitude at the outset. At the end of the day, I needed the flexibility to create and track events on the fly. Their interface was very easy to configure and the out-of-the box features were exactly what I needed. I especially loved their Path analysis (shown below).

Pendo Path Analysis.jpg

It shows me where users came from before they took a specific action. For example, I wanted to know where users were before they clicked on our Help icon to get to our knowledge base. By looking at this Path I could identify which pages needed attention. It turns out many users were clicking the Help icon when they were in our settings pages. This was a good indication to me that these pages needed some attention. It was a good starting point to start digging further.

Ultimately, there are many great tools out there and as they continue to improve I’m sure the market will become more and more competitive. Take some time to explore these different tools and choose the right software that suits your business needs.

Getting Customer Feedback

This past winter my product team and I embarked on a customer interview ‘world’ tour… And by world tour I mean we went to Europe. Our goal was to talk to as many customers and non-customers as possible. The business was challenging us to come up with a compelling roadmap and to identify problems that needed to be solved. We had just hired a new Head of Product and he wanted to level set, to go back to the beginning and truly understand why this business existed. He felt strongly that in-person interviews were worth the extra expense and made the decision to have us travel to our users versus doing video calls. It did take some convincing of our CFO, but ultimately, our new boss felt it was the only way we were going to get quality interviews (at least to start), by being in the space with our users, seeing what they saw on a daily basis and stepping through the looking glass.

Without a doubt, the most important playbook for a Product Manager to have is a deep understanding of who their users actually are. Often, I find that we make assumptions, sitting behind our cozy, familiar desks, far far away from our actual users. It’s easy to forget the WHO and the WHY when most of our days are spent figuring out the WHAT.

Before we packed our bags for Europe, we went through a training on how to conduct in-person interviews. We practiced on ourselves and each other. We even had to record ourselves as the interviewer; playing back our ‘game tapes’ to understand what we needed to improve. It was worth the effort, but also very cringe worthy. Listening to yourself and seeing how you respond is always hard at first. I learned a lot of what not to do and found that I had a tendency to lead the user in their responses. I nodded my head a lot and agreed with what they had to say. I wanted to encourage them, thinking this would help me get better responses. But in actuality, it was biasing them. Causing them to respond in a more positive way than if I had just remained silent. I’m still working on this one, but I notice it earlier in my conversations and am able to course correct as a result.  

 Here are 11 key takeaways from that user interview training that I think can apply to all interview situations: 

  1. Start with approximately 4 user interviews, try for 6-8. You don’t need to interview tons and tons of users off the bat. Start small. Be realistic. Patterns will emerge sooner than you think.

  2. Keep the interview 1:1 (if possible). When too many people are in the room at once, you will find that one personality will dominate the conversation. Having a one-on-one conversation is the best way to dig into real problems.

  3. Have someone take notes. Their job is just to listen and document what happened. I found these notes to be vital, especially after conducting more than one interview in a day. Everything can easily start to blur together. If your note taker can jot down interesting quotes, that can be very handy in the future.

  4. Record sessions (if you can). Make sure to ask for the interviewees permission before doing so. Try to make the recording inconspicuous. Seeing yourself on camera can be very distracting.

  5. Take a minute to warm up your interviewee. A casual conversation will help them relax. Most of us are not used to being interviewed and it can be a little unnerving at first.

  6. Ask them to perform tasks versus answering questions directly. If you can, try to get them to show you instead of telling you. A lot of times people explain things differently than how they actually perform them.

  7. Don’t explain what they are doing wrong. This is a hard one for me. My instinct is to want to help, but by correcting them you are taking over their authority. It can also derail the conversation. Stay quiet in the moment. You can always follow up later.

  8. Try not to talk too much… like as little as possible. Your job as the interviewer is to LISTEN to them and make them feel heard. If you talk too much, you won’t get a lot out of the conversation.

  9. Keep a neutral response. As I eluded to earlier, nodding and agreeing is actually the complete opposite of what you should do. I know it’s hard, but just be Switzerland. You are neither happy nor sad. You are the face of neutrality. Good luck – this is a tricky one!

  10. People won’t be mean to your face (for the most part). Remember that ultimately there will be some bias in users’ responses. If you start to sense some resistance, ask them if they are comfortable talking about the subject. If they are not, move on. This is not the time to pry or make them feel uncomfortable.

  11. Echoing back instead of just asking why. This is a powerful technique to ensure you understand a user. You will be surprised at how often you misinterpret what they are saying. Clarifying is always helpful.

The next thing our training course helped us prepare for was how to ask good questions. More than just setting the stage is ensuring that you are not influencing the person you are interviewing. Many times we find that research results can just be a validation of our assumptions. Interviewers tend to ask questions that lead to answers they subconsciously (or consciously) are seeking.

Here are 5 tips for how to ask good questions:

  1. Answer questions with questions. If the user asks you if they are doing something right don’t respond with a yes or no answer. Instead ask “What would you expect to happen?” or “What thoughts were going through your head when you did that?”

  2. Ask open ended questions. They usually begin with:

    1. Why

    2. How

    3. What

  3. Avoid close ended questions. They often start with:

    1. Do

    2. Is

    3. Are

    4. Would

  4. Write intended questions in advance. We created a template for each type of user persona we interviewed. Note this is not a script, but rather a guideline for asking similar and targeted questions to each user.

  5. Embrace awkward silence. Long pauses are OKAY. This one was so hard for me. If you try to fill space, like I did at first, you actually cut short a lot of interesting conversations. If you leave them open, it will force the user to think more deeply and respond further. Yes, it’s awkward, but you will be shocked at what responses you get after.

 In the end, we interviewed over 60 customers in just 3 months. I came out of this experience with a real sense of WHO are users were and what challenges they faced on a daily basis. It was such an eye-opening experience. I felt the passion many of our users had for what they do and also empathized with the struggles they experienced.  

At the end of every interview we conducted, we did a quick 5 minute de-brief between the interviewer and the note taker. We jotted down a few bullets on what we found most interesting about the conversation. This was so HELPFUL when we did our analysis later on. It made referencing back quick and easy, and surfaced the key points of each conversation. I’ll save our analysis project for another blog post, but I hope you are able to take some of these tips and use them in your next interview. You might just learn something new. Good luck!

Product Usability and Engagement Metrics

Overview

User metrics give us clear insight into the successes and failures of our product(s). They provide a means for us to hold ourselves accountable to our objectives. If you’ve never worked with user event tracking tools before then this article is a great place to get you started. In the past, I have rolled out event tracking tools at 2 different companies and trialed 4 different products. They are wonderful to have, but take time and dedication to setup properly.

Why are they important? 

What you don’t measure, you can’t improve.

  • Validate product decisions with data, not your gut.

  • Gain insight into user behavior. These insights in turn drive specific in-app and marketing improvements.

  • Better prioritize feature improvements and bug fixes. Tracking user behavior can indicate where problem areas are.

The Difference between Usability and Engagement

Usability metrics focus on how easily people can accomplish what they’ve set out to do.

Engagement metrics focus on how much time people spend interacting with your product (best measured in a flow state) or how good they feel about it (qualitative). 

Metrics to Consider

  1. Number of Sessions per User (per day/per week/per month)

    1. Measures: How often users log in (frequency)

      1. Tip: Avoid averages and take medians over means (less sensitive to outliers)

      2. Why? Once a target has been established we can use the median to identify customers that may be likely to churn, need assistance, or are having issues (proactive versus reactive).

  2. Session Duration

    1. Measures: How much time users spent interacting with the product or in a particular flow state

      1. Tip: Use medians instead of means (less sensitive to outliers)

      2. Why? Identify at risk customers and test feature usability if measuring a flow state. Establish a benchmark and analyze usage over time to identify trends. 

  3. Number of Key User Interactions per Session

    1. Select user interactions that matter the most (e.g. clicks on “Publish”)

      1. Why? Validate assumptions, test “important” features and identify those that could be eliminated or simplified

      2. Tip: Look at difference cohorts of users (by user role, permission set, etc…)

      3. Idea: Compare the differences in the metric between churned and retained customers within 30 days

  4. Feature Usage

    1. Measures: How frequently a feature is used compared to assumptions or other notable features

      1. Tip: Be specific in the expected flow or path a user is expected to take. Identify bottlenecks, bounce rates, and alternative paths

      2. Why? Use to refine/improve usability or remove features

  5. Task Success Rate

    1. Continuation of Feature Usage

    2. Measure: Time on task vs. expected

      1. Tip: Look at back-button usage and error rate (instead of bounce rate)

      2. Why? Use to refine/improve usability or remove features

Next Steps

  1. Identify the top 3-5 experiences in your application (what is it you expect a user to accomplish?)

    1. Breakdown by key interactions

  2. Evaluate tracking tools (see my next article for a full run down of the tools I’ve personally evaluated)

  3. Implement tracking tool

    1. Create a naming convention for each event - very important!

    2. I like to use the following - Page: Feature: Action

  4. Set usage metric goals for all new features being rolled out

  5. Measure

  6. Refine

The Best Approach to the MVP

Defining the MVP is always harder than it looks. Cutting features you are excited about can feel like defeat, but at the end of the day you have to remember the bigger picture. Recently someone shared with me this great visual of how to think about the MVP. Whenever I am confronted with the decision to cut scope or reduce features I think about this evolution.