Alpha Co-Creation Loop
Pick the users who are already pushing the boundary, build with them in a shared channel, ship only when they're thrilled.
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 95%
Stripe's default product development method: rather than researching a market and launching to it, you select a small set of existing users who are already straining against the limits of what you offer, bring them inside a shared development loop, show them the product on a regular cadence, and only graduate to a broader audience once that alpha group is genuinely delighted. The selection of the alpha group is the strategic act — they are chosen because their needs ladder up to the strategy, not because they're loud.
Origin
Described by David Singleton as 'just how we build product at Stripe.' The canonical case is Stripe Billing, co-created with Figma, Slack and others; Stripe Connect emerged the same way from working with Lyft and Shopify.
Core principles
- 01Build the next layer where user need meets adjacency to what you already do — that confluence is the invitation to build.
- 02The right alpha users are the ones already pushing the boundaries of what's possible with your current product.
- 03Choose alpha users whose needs ladder up to your strategy, or you will optimise for the wrong future.
- 04Delight of the alpha group is the ship gate — not a date, not a feature checklist.
- 05The loop only works if you can get changes into their hands fast; speed of delivery is a prerequisite, not a nice-to-have.
- 06Engineers who run this loop develop most of the attributes of a PM, which is why Stripe went years with almost no PMs.
How to run it
- 1
Find the adjacency
Look at what your users get from your product today, then at the problems adjacent to the ones you already solve, which they are pouring their own energy into and feel they shouldn't have to. That confluence of need and adjacency is the signal to build.
Watch out Adjacency without need produces a roadmap-shaped product nobody asked for.
- 2
Select the alpha cohort by strategic fit
Pick existing users who are visibly straining against your current limits and whose shape you expect to become common. For Billing these were Figma and Slack — payments users with subscription business models, at a time when many more such companies were clearly coming.
Pro tip Ask: if we perfectly serve this cohort, does the resulting product serve the future market? If not, pick a different cohort.
Watch out Picking the loudest or biggest users instead of the most strategically representative ones will bend the product toward one-off requirements.
- 3
Get to know the individuals, not the logos
The team personally understood the individuals actually operating those systems at those companies — their challenges, hopes and dreams for the future. The relationship is with people, not accounts.
- 4
Open a shared, always-on channel
Stripe ran shared Slack channels with the alpha companies. This collapses the latency between a user noticing something and the team hearing about it.
Watch out A quarterly advisory board is not this. The whole point is continuous, low-ceremony contact.
- 5
Show product on a regular cadence and take the feedback back into the loop
Demo the actual product to them frequently and feed their reactions straight back into development. Continuous deployment matters here: feedback in the morning can become a change in their hands by the end of the day.
Pro tip If you have the loop but can't get changes into users' hands rapidly, fix the delivery pipeline — that is the bottleneck on learning.
- 6
Gate the broad launch on alpha delight
Only when the original alpha group is 'super, super happy' with the product do you consider it ready for a broader audience.
Watch out Launching before the alpha group is delighted means launching a product whose sharpest users are still unhappy — the broad audience will be worse off, not better.
In the wild
Companies like Figma and Slack were already using Stripe for payments but had subscription business models and were pushing the boundaries of what was possible. Stripe brought them into a shared development loop with shared Slack channels, showed them product regularly, and got their feedback continuously.
→ Billing was only taken to a broader audience once the alpha group was super happy; it became Stripe's revenue and financial automation suite, later extended with complex multi-year deal modelling built with Atlassian and Cloudflare and shipped to general availability.
Lyft and Shopify were using Stripe for pay-ins in the early days. Working closely with them, Stripe realised these companies were operating multi-sided marketplaces with an enormous amount of regulatory and money-movement heavy lifting.
→ That observation became Stripe Connect, the product for platforms and marketplaces — a valuable space Stripe reached by solving one cohort's adjacent problem rather than by market analysis.
Common mistakes
Having no feedback loop at all
Singleton says it is remarkable how frequently a real user feedback loop simply does not exist in product development cycles. If you don't have one, creating it is the single highest-leverage move available.
Having a loop that is too slow to close
A feedback loop where the idea-to-user-feedback cycle takes weeks provides too few learnings per unit time. The fix is in the delivery infrastructure, not the research process.
Choosing alpha users for convenience
Alpha users who are easy to reach but not straining against your limits will validate the product you already have instead of pulling you toward the one you need.
Is it for you?
Best for
B2B and infrastructure product teams with an existing user base, deciding what to build next and who to build it with.
Not ideal for
Consumer products where no single user cohort is representative, or true zero-to-one products with no existing users to co-create with.
From the transcript
“The way we think about product development, it's it really is to find the correct set of early users to kind of co-create the product…”
“And only when that original kind of alpha group was super super happy with the product did we then think it might be ready to…”
“If you have a mechanism to listen to users, get something in their hands quite quickly, and then get their feedback on it to run…”
From the episode
Building a culture of excellence
David Singleton (CTO of Stripe)