Thread by @johncutlefish

the “real world” of continuous improvement ... from a week of chats w/external teams (mostly non-customers, non-prospects) at @Amplitude_HQ

1/10 after 3m of hard work, a team resolved a major issue with CI/CD and implemented some basic feature flagging.

A step forward...

2/10 a “task force” of 5ppl (cross-functional) implemented lightweight governance for instrumenting new features/workflows across 15+ teams. Took 1m. Where before there was a mess of events, now there was a common approach and language across teams.

A step forward...

3/10 a successful experiment doing a weekly, fixed-time usability testing session. After six weeks, the unanimous feedback — across engineers, designers, product managers — was to keep the session. The new challenge...more #ux researchers!

A step forward...

4/10 ...for the first time, a team experimented with a “pivot/proceed” point for a new effort instead of some random definition of done. Now...it got messy. And they learned something new about how they measured something, but that was part of the learning.

A step forward...

5/10...Instead of having a “tech” roadmap and a customer-facing roadmap, team experimented w/putting it all on one board. 1 month experiment of doing apples - apples prioritization.

Learning? Tech leads needed support and product training.But they kept it.

A step forward...

6/10...A legitimate “win”, but only after five attempts to solve a nagging onboarding challenge. The solution? The super lightweight FAQ performed way better than fancy modals and sequences....

A step forward...

7/10...After 7 months of looking, a team finally found a CPO who meshed with their product culture.

A step forward...

8/10...A 2m pilot of Data School, with experts in both qualitative & quantitative research/analysis.

Learning? Though lots of people applied & attended, a smaller % actually seemed interest in carrying the skills back to their teams. Pivot to advocate program

A step forward...

9/10...After 9 months, someone left their company after really giving it a healthy try.

A step forward...

10/10...A team experimented with a fully cross-functional “discovery” track, allocating 30% time for the team while other work was ongoing.

Learning? The multi-tasking hit was too much. But they “loved” the discovery activities.

A step forward... (one more...)

Why share all this? This is the “real world”. Product development is hard.

Imagine a spectrum between “that will NEVER work here” and “JUST DO THIS” ... in the middle is where teams actually (hopefully) exist. Not afraid. And not copy-pasting things context-free.