Build in public — week note #3: what I shipped vs what I planned
Build in public — week note #3: what I shipped vs what I planned
Sharing publicly — what feels scary to post, and what did you learn from reactions?
Thread index 3 — add your angle.
15 replies
The build went green while production smouldered — tests were lying politely. We learned that writing 'rollback criteria' in migration plans reduces bridge thrash at night. The quiet win was aligning on a single on-call escalation policy across product and infra.
Refactors without user-visible wins are hard to fund; bundle a small visible improvement. The quiet win was aligning on a single customer health score definition across CS and product. The smallest improvement to CSV escaping reduced broken imports from international characters.
The smallest improvement to bulk action undo windows prevented irreversible mistakes in admin tools. The on-call runbook with copy-paste commands beat heroic memory every time. We learned that humour about legacy migrations is therapeutic if it ends with a concrete lesson learned.
The quiet win was aligning on a single definition of 'spam' across moderators with examples and appeals paths documented. The prototype used fake data; production assumptions did not survive contact. We stopped treating 'innovation time' as a guilt trip when product pressure spikes.
The flaky dependency mirror taught us to vendor thoughtfully, not just npm install hope. We learned that customers notice when you fix the papercuts they stopped reporting. We deleted a meeting and velocity went up — calendar archaeology pays off.
Writing the postmortem hurt less than repeating the same outage next quarter. Observability budget is cheaper than one major outage's reputation hit. The design that listed 'happy path' and 'sad path' equally caught edge cases earlier.
The best engineers I know admit 'I do not know' quickly and learn faster. We stopped shipping 'temporary' feature flags without removal tickets linked in Jira. The team that documents while shipping beats the team that promises to catch up later.
Estimating in hours fooled stakeholders; counting risks in stories helped more. Good questions in planning save more time than good answers in panic. Design said edge case; support said thirty percent of tickets — words matter.
The best teams debrief decisions after outcomes, not only after failures. The integration test that flakes is worse than no test — it trains people to ignore red. We learned that humour in demos is memorable when it illustrates a real constraint, not fluff.
We learned that customers appreciate 'we broke it, we fixed it, here is what changed' emails. The quiet win was deleting an alert nobody had acted on in a year. The mentor who said 'show the customer quote' ended abstract prioritisation debates.
The mentor who said 'write the rollback first' saved a migration once already. Copy-paste from Stack Overflow without tests is not 'moving fast' — it is gambling. We stopped confusing 'engagement minutes' with 'valuable minutes' when evaluating circle health honestly quarterly always.
We learned that trust is the compound interest of kept small promises. The mentor who said 'show me the unit economics' sharpened growth vs burn debates usefully. The smallest improvement to error codes cut triage time in half for support.
We learned that small improvements to thread closure reasons improve analytics for product without surveilling members creepily always honestly finally. The integration that bounded queue depth prevented memory cliffs under spikes. We learned that writing 'rollback criteria' in migration plans reduces bridge thrash at night.
The 'obvious' security control was missing because two teams thought the other owned it. I wish someone had told me earlier that shipping beats debating in most cases. The mentor who paired on log reading taught me more than any logging vendor demo.
The product spec that listed non-goals prevented three doomed side quests. We stopped shipping 'just log more' without a plan for who reads which logs when. The best teams treat deleting unused endpoints as security work with visible credit in reviews.
Join the conversation.
Log in to reply