Held on 25 August 2022
Don't Test Until It's Live
We all hate the software errors that chew up our airplane tickets or delete our carefully crafted documents. "Didn't anyone test this thing?" we scream, frustrated that anyone would make us buy such poorly crafted services. But in fact, sometimes the best testers are in fact our users, and we'd do much better to learn directly from them than to try "engineering in" quality.
I have direct experience with "bug-friendly" environments where the very successful philosophy was that since bugs are inevitable, we should control and detect them in the wild, without holding back the development of new features for artificial "user acceptance testing" or similar lengthy interventions. And this isn't just in low-impact domains like e-commerce; high-frequency traders regularly and successfully "spike and stabilise" new systems, and I've even seen biotech and insurance companies drop complex compliance procedures, replacing them with a focus on error detection and recovery.
Join me on this free livestream to discuss:
- Why "beta" is a bad word, and alternatives that keep users safe while letting some of them try out something new safely.
- How to go way beyond A/B tests to verify multiple hypotheses or try a new feature on a controlled sample of real users.
- The "big red button" that mitigates fears and can enable courageous releases.
Sign up to get a reminder and receive bonus material.
