On passing an impasse at Amazon
When Design and Product Management are at each other’s throats, a mild provocation can get the process back on track
Audible (a subsidiary of Amazon) has long been the industry leader in audiobooks, but when I began there in early 2015, they were hard at work on a new podcast product called Channels.
And if 2015 sounds a bit late to be arriving to the podcasting party…well, Channels came with some baggage:
some baggage
Originally conceived as a standalone app, at some point it was decided that Channels would be more likely to find success within the main Audible app. Which sounded reasonable enough, though the reality proved a tad trickier.
Rather than having an entire app to fill, the new direction now required Channels to fit within a single tab. The product manager didn’t want to abandon any of the planned features, however, so his solution was an unlimited number of tabs within the tab:
tabs as far as the eye can see: Recent Listens, Noteworthy, Categories, etc. etc. etc.
A clever approach, to be sure, but the design team had some concerns about the proposed structure:
1. It created a fundamental imbalance, as there was no precedent for such complex functionality anywhere else in the app
2. It asked a lot of new users, and early testing indicated that they were unwilling to get past the learning curve
In short, it was an app hidden within an app — and such “Russian doll” architecture almost always signals a mistake.
. . .
At the time, Audible had a dozen Agile teams each devoted to a different platform, product, or feature. In my role as experience architect, my job was to develop standards and then work with the various teams to ensure that what they produced was consistent with the larger vision.
As Channels was the most high-profile and high-risk project, I kept close tabs on it for nine months, watching as the relationship between Design and Product deteriorated to a standstill. Design said the UX was broken and Product said every feature was needed — and despite several attempts at escalation, no one had budged.
Sensing that something drastic was needed, I took two days and designed a radical alternative. Instead of asking the customer to configure the experience, I created logic to let it work instantly:
math & manners
Instead of unlimited tabs, I allowed just two — one for discovering and one for listening:
symmetry & balance
Instead of a lengthy onboarding process, I used notifications and error messages to educate the customer as they explored more deeply:
“need-to-know” architecture
My counterargument quickly made the rounds and soon the VP of Product was asking the VP of Design if my suggestion could be used. But he knew as well as anyone else that it was too late in the game to hit the reset button.
What did reset, though, was the conversation. Everyone started talking and thinking again, and that was my only goal. If I had a point beyond that, it was merely to show that there’s never a single answer — the more complex the problem, the more likely that it can be solved using any number of methods.
And one of the best methods, of course, is compromise. Looking at the Audible app in 2018, Channels is still in its own tab, but now it’s been pared down to just those two key areas I suggested — discovering and listening:
Audible app for iOS (circa 2018)