Beyond Customer Interviews: From Post-Its to Product

Now that you’ve nailed conducting killer customer interviews and are getting good customer stories; what’s next? How do we take a collection of conversation notes and turn them into concepts for awesome products?

Let’s take a look at one way through the product design process.

One of our go-tos for compiling and analyzing feedback is the affinity map. To affinity map your feedback, take all of the quotes and observations from your interview, write them on Post-its, and stick them to the wall. Once they’re all up, move the stickies around to group “similar” things together.

The groups that emerge from this exercise will lead to a number of insights and questions. For example, “It looks like most customers shopping in groups took multiple sizes into the dressing rooms at once. Lone shoppers didn’t; I wonder if social shoppers optimize trips to the dressing room while less frequent shoppers don’t. If so, I wonder why?”

Repeat this several times, remapping the sticky notes according to different similarities. Though it may seem redundant, forcing the team to do this one more time than feels reasonable can lead to great insights. Also, don’t shy away from seemingly strange groupings. Strange can be important.

Some of the insights you gather will help define customer segments – classes of customers that behave in similar ways, have similar needs, and share demographics. Some will help identify problems or needs of the customers.

So what do you do with a bunch of customer segments and insights about their needs? A great tool we often use is the problem statement:

“CUSTOMER SEGMENT needs a better way to VALUE PROP because PROBLEM/NEED”

It’s important to come up with several of these statements before selecting one. While it’s easier to rush through the process and settle on the first thing you come up with, this proves to be far less fruitful in the long run.

Even more importantly, make sure you don’t embed your ideas for the solution into the problem statement. This often results in a problem statement with only one possible solution. For example, “College students need a better way to order pizza because they can’t order pizza from their mobile devices.” A better alternative is, “College students need a better way to order pizza because they often want one while out dancing, where it’s too loud for phone calls.”

To identify a good problem statement, start with a quick brainstorm of solution ideas and questions. A good problem statement makes it really easy to come up with ideas. Don’t worry about the quality, just make sure you can come up with many ideas and questions.

Problem statements that are too narrow will only allow for a few ideas that fit. Those that are too broad lack structure and lead to blank page syndrome.

With a problem statement selected and a long list of (mostly silly, strange, or outright bad) concepts and questions, it’s time to pick a few favorites. You can do this by dot voting on the brainstorm stickies, giving everyone on the team three votes and stack ranking the results based on votes.

Great work. Now that you have some concepts and questions in hand it’s time to start Customer Validation.

More on that later.

How To Destroy a Month of Work in a Minute

A month of coding down the drain.

“What are the top features you’d look for in a Pinterest analytics product?” we asked the client, prepared to hear an agenda for our demo.

We expected to pull back the curtain on a solution for exactly those problems, and awe them with our intuition and customer insight.

A minute later we realized we’d failed.

Our client’s answer to our question was not what we’d anticipated. We didn’t support any of the use cases they enumerated. Every single feature they mentioned focused on tracking activity across all Pinterest users for their ecommerce site. Our entire product focused on engagement around their Pinterest account. We’d completely missed the mark.

Rewind a month.

We’d just finished meeting with a number of friends at social media agencies, where many had expressed a need for analytics for Pinterest. Pinterest was new; it didn’t yet support any tracking or work with any of the big analytics packages. Also, the existing products were overly complicated for the social platforms they did support. They wanted something simple, something they could use to show their clients they were doing a good job.

Perfect. We could easily solve this for them. We started work on an MVP, confident that the Pinterest API could easily cover this use case.

Except it wasn’t easy.

As we moved higher within the organizations, engaging with financial buyers, things became more complicated. At the time, social media was seen as a bit of a black art. A lot of social folks operated on instinct rather than data. Buyers didn’t want to pay for something that risked showing their efforts in anything but the best possible light.

No problem. We’d switch customers, focusing on their clients instead. These clients, big ecommerce brands with a strong social presence, definitely wanted to know social engagement on their accounts, right?

We emailed a few friends in social media at some big brands and asked them if analytics for Pinterest were important to them. They said yes. Slam dunk. We spent the rest of the month frantically finishing the MVP to demo.

Which brings us back to the client demo. What did we do wrong?

We switched customer segments without properly re-examining the value proposition. While both customer segments wanted Pinterest analytics, their understanding of “Pinterest analytics” varied greatly. To the agency folks, “Pinterest analytics” meant social engagement data from campaigns run on Pinterests site; new boards, successful repins, etc. Conversely, to brands “Pinterest analytics” meant data from engagement with the brand across all of Pinterest. Essentially, they wanted to know how many times images from their website were pinned. The minor difference of definition resulted in a drastically different feature set.

We’d built a product for our first customer segment but were trying to sell it to our second. And we didn’t even realize a difference until our first sales pitch.

We’ve learned our lesson the hard way; we now always vet our assumptions by asking customers to explain the value proposition back to us, in detail. Remember, while everyone may claim to go to the gym to get “in shape”, depending on who you ask, “in shape” may mean “ready for bikini season” or it may mean “strong enough to lift a car”.

The old annoying cliché about assumptions is half right. I certainly felt like a bit of an ass leaving that meeting.