Life in Beta
Test small changes, collect feedback, and keep only what works
- Difficulty
- Easy
- Time to result
- ~ongoing to results
- Steps
- 5
- Confidence
- 96%
Perkins describes Bonzo as always being in beta: the team continually tests AI ideas and other features, lets beta users interact with them, and discards the large majority that do not work. The mechanism is a progression from micro experiments to macro moves. An idea earns further investment only after use and feedback reveal that it helps. Perkins then generalizes the same loop to life, relationships, family, and career: regard the present version as changeable, try deliberate adjustments, observe their effects, and keep becoming better. The model normalizes rejection as part of innovation. A high discard rate is not evidence that the process failed; it is evidence that testing prevented weak ideas from reaching the market.
Origin
Perkins connected Bonzo's continuous R&D and beta-user feedback process to a wider lesson about improving one's life. Extracted from Coffeez for Closers.
Core principles
- 01Treat the current version as provisional
- 02Use small tests to inform larger moves
- 03Expect most experiments not to survive
- 04Let feedback stop weak ideas
- 05Apply iteration beyond products
How to run it
- 1
Frame an uncertainty
Choose an idea whose value is not yet proven. State what improvement it is supposed to create.
Pro tip Prefer a narrow friction point over a broad innovation slogan.
- 2
Run a micro test
Build or try the smallest version that can expose whether the idea helps. Avoid treating the test as a finished commitment.
Watch out Do not market an experiment as a proven capability.
- 3
Get real feedback
Let relevant beta users experience the change and report what happens. Look for evidence that the intended outcome actually occurs.
Pro tip Observe behavior as well as stated preference.
- 4
Discard or advance
Stop ideas that do not work and advance the minority that demonstrate value. Use successful micro steps to justify larger steps.
Pro tip Keep the rejection threshold high enough that weak ideas do not drift into production.
Watch out Do not interpret sunk effort as evidence.
- 5
Repeat across domains
Apply the same test-and-learn stance to business, career, relationships, and personal habits where appropriate.
Watch out People and relationships require care; they are not disposable product experiments.
In the wild
Perkins says Bonzo's R&D department continually tests AI and other feature ideas with beta users. He estimates that 90% of what the team tests never reaches the market because feedback shows it will not work.
→ Weak ideas are stopped before release while the useful minority can support larger product moves.
Common mistakes
Falling in love with every test
An experiment is a question, not a commitment; many ideas should be rejected after feedback.
Taking a macro step first
Large investment before a small evidence-producing test reverses the sequence Perkins recommends.
Is it for you?
Best for
It is best for product teams and individuals operating under uncertainty.
Not ideal for
It is not ideal for decisions where uncontrolled testing would create unacceptable safety or compliance risk.
From the transcript
“you take baby steps to take macro steps right take micro steps and take macro steps”
“90% of the stuff that we test doesn't even make it to Market”
“our life should be in in in beta all the time we should constantly be testing ourselves”
From the episode
Conversations at Scale ft. Jason Perkins
Jason Perkins