A Case in Product Discovery

A real world (fictional) example for Product Discovery

Product Fundas
Bootcamp

--

The Case

Jay is a Senior Product Manager at Amazon responsible for Amazon Music

That is a Friday. Jay is at his desk. Alone. Engrossed.

Jay’s calendar notification pops up. It reminds him that it is time to sequester himself to some solo, heads-down thinking on ideas to make Amazon Music better.

Jay eagerly looks forward to this alone time every week. He is already on the task. He closes the notification. He sets his iPhone and Mac to ‘Do Not Disturb’, and soon enough, he is back into deep rumination.

Jay’s plan is to deep-dive into consumer feedback, which he does religiously every month. He has a simple process: review the output generated by an AI engine that performs some heavy-duty analytics on raw ratings and feedback data.

Jay furrows his eyebrows zooming in on one feedback. It says:

Amazon Music needs to offer social networking ability for people to connect with and interact around an album, a song, an artist, a podcast, a topic or anything related to music or podcast.

Jay quickly opens the reports for the last three months. He observes that this is a consistent feedback from many users.

What should Jay do now? As the Product Manager, how should Jay proceed?

React vs. Respond

In my opinion, there are only two ways forward — react or respond. There’s a difference.

  • Reaction is typically quick, without much contemplation, and mostly driven by one’s beliefs and biases. On the spectrum of all possible options available to deal with the situation, reaction typically are the extremes.
  • Response is typically calm, well-thought-through, and driven by facts and data

If Jay were to react, the two extreme positions could be:

  1. Jay agrees, “It is a great idea to build social features into Amazon Music! Why didn’t I think of it earlier?” Jay proceeds to imagine some features, maybe gets inspired by other social platforms. He then documents the PRD and sends it up the flagpole for approval.
  2. Jay dumps the idea saying, “Amazon does not own a social platform. Amazon Music is certainly not a social platform. It is a music streaming service. There are workarounds on the platform that allows the user to network socially. It is absurd to turn Amazon Music into another social network.”

If Jay were to respond, then there is really just one way forward —begin with problem discovery.

By the way, I’m not passing any judgment on react vs. respond. In some cases, reacting may well be the best way forward. In other cases, responding could be the best foot forward.

In this case, I believe it is clear as daylight that Jay would be much better off if he chooses to respond by kicking-off a new problem discovery thread, because reacting would end up being just knee-jerk.

Product Discovery — a quick recap

Problem Discovery involves the following steps, at a high level:

Identify problem & User segments -> Discover root causes & motivations -> Estimate user segment size -> Calibrate problem freq. & severity -> Validate vision alignment -> Determine relevant metrics -> Brainstorm solution ideas -> Validate ideas

All of the above, before the solution is built. No scratch that. It is before we even prepare to build anything.

Product Discovery — the Amazon Music Case

Step 1: Problem — User Segment/s — Motivation — Engagement

Problem:

The word problem is usually used to refer to symptoms, like the lack of something in the product, say an ability, a feature, usability and so on.

In Amazon Music’s case, the problem clearly is the lack of social features. The abilities such as like/dislike a song, share a link via email, Facebook or Twitter, or copy-paste the link on any other platform, hardly count as social features.

User Segment:

The idea here is to construct the persona of the users who are sorely missing the social features.

It is not straightforward to pinpoint the specific user segment/s. There are several ways to collect data to construct a user persona and box the user segment/s. For example,

  1. Users who use the like/dislike feature, with a hypothesis that these users most likely want the social features.
  2. Users who use the share feature, with a hypothesis that those who share are even more likely to want the social features.
  3. Users who provided feedback (provided it is a statistically significant number)
  4. In-app survey (or external survey) — If none of the above methods work, construct a survey based on the above data and target it to a larger set of users.

Motivation:

Now that we have the user segment/s nailed down, we begin with the why questions. The purpose is to discover the root causes of the problem, as any solution needs to address the root causes or the users’ motivations:

  • Why do these users need social features within the app?
  • What satisfaction might they derive from in-app social features that is not gained by simply copying a link and sharing it on other platforms?

Oftentimes, users mask their motivations. They disguise their imagination of the solution as their needs.

Almost nobody says, “Hey, here are my motivations. How can you help meet them?” Users simply say, “I need social features.” As a Product Manager, it is expected of us to investigate why (underlying motivations), and then solve for them.

Breadth-Frequency-Severity:

Finally, a Product Manager must consider the numbers: how many users are expected to use it? How often will they use it?How sorely are they missing it today?

Step 1 ends with a decision on whether it is worth investing in building these features.

Step 2: Vision Alignment & Metrics

Vision Alignment:

The next step is to validate if solving for the root causes, or users’ motivations align with the company’s and product’s vision.

Amazon’s vision is to “be the most customer-centric company on earth”

Amazon Music’s vision is “to be the best value music service for Amazon customers, not a me‐too on‐demand streaming service”

Would adding social features continue to keep Amazon and Amazon Music aligned to their respective vision? Some follow up questions would be:

  • Does customer-centric mean that Amazon builds every feature that the customers request?
  • Would social features count as the best value music service for customers?
  • Would not adding social features make Amazon Music a ‘me-too’ service (because no one else does)? If yes, should they build the social features?

Metrics:

Every Product Manager tracks metrics: NSM, Revenue, Usage, Usability, NPS, and so on.

Metrics are one of the key considerations for decision-making. The simple question would be: which Amazon Music’s metrics would the social features help move, and to what degree? Are these the primary metrics, like NSM?

For example, suppose the NSM for Amazon Music is Cumulative Streaming Hours. Would social features help increase CSH? It could if the social features are built in such a way that they become accessible while streaming is in progress. Otherwise, they may not.

Step 3: Ideas for Solution

The Product Manager has two decision-making points, one after Step 1, and the second after Step 2.

If the decision is a ‘go’ at both these points, that is when the ‘how might we’ questions begin:

  • How might we get users to interact with each other?
  • How might we get users to post their thoughts about a song, an album, an artist, a playlist, etc., and share a link to it while they are at it?
  • How might we get users to like someone’s post or add their comments to it?

Questions like these help generate ideas for solutions. Some basic UX design will help test out these ideas. Maybe we iterate a while with different solution ideas.

Mind you, there is zero implementation effort invested. We’re still within the boundaries of Product Discovery.

Product Discovery helps the Product Manager to develop a strong business case for an idea for product improvement and run it up the flag pole for approvals.

Conclusion

To conclude, it takes some intense research, analysis, and decision-making before choosing to make any significant improvement to a product.

The challenge is to engage in such an intense analysis while simultaneously preventing analysis paralysis.

That’s the post, my friends. Thanks for reading till the end.

If you have some views about this piece, please feel free to add your comments.

If you liked this piece, go ahead and share it with your network. And of course, tap the clap icon below and show me your appreciation. Follow Product Fundas.

--

--

Stories, anecdotes and learnings on Product Management & Design Thinking. No BS. Only actionable ideas. Tweets @productfundas