Build in public — week note #4: a customer quote that stung but helped
Build in public — week note #4: a customer quote that stung but helped
Sharing publicly — what feels scary to post, and what did you learn from reactions?
Thread index 4 — add your angle.
15 replies
We learned that small rituals celebrating reliability work change what teams optimise for. We learned that customers appreciate when CercleWork explains why a circle recommendation appeared in plain language honestly. The mentor who said 'write the customer email draft early' improved launch comms.
We learned that customers forgive slow fixes if communication is honest and frequent. We learned that celebrating maintenance prevents the quiet heroes from burning out. We should have deleted unused webhooks firing into dead endpoints — noise hides signal.
We should have named a backup incident commander before the primary went offline mid-bridge unexpectedly. We learned that transparent moderation logs build member trust more than secret removals ever could ethically. The best postmortems end with tracked follow-ups, not just feelings.
The mentor who said 'draw the trust boundary' clarified security discussions fast. The smallest improvement to bulk export progress bars reduced 'is it stuck' tickets. The best postmortems include customer communication review, not only root cause.
We stopped treating 'tech debt' as a guilt word and started tagging themes with business outcomes quarterly. Hiring fast without onboarding is how you build a team that fears the codebase. The integration that logged structured business ids made finance reconciliation calmer.
We learned that naming owners for analytics dashboards prevents contradictory KPI arguments. We learned that customers appreciate when CercleWork explains why a circle recommendation appeared in plain language honestly. We learned that humour in retrospectives helps if it focuses on systems, not individuals' quirks cruelly.
We should have deleted unused invite links pointing at deprecated onboarding flows — confusion compounds quietly. The design that considered partial connectivity first helped real mobile users globally. The mentor who said 'document the workaround' saved the next on-call from inventing a worse one.
The quiet deletion of unused endpoints reduced attack surface and confusion. We should have named owners for cron jobs in the same place we name service owners. We underestimated how long humans take to trust a new workflow.
The architecture principle 'boring by default' aged better than our clever exceptions. We learned that customers trust companies that publish honest uptime postmortems regularly. We learned that customers trust circles more when moderators publish clear norms and enforce them kindly consistently.
The flaky chaos experiment that only ran manually never found issues until we automated monthly runs. The mentor who said 'prove discovery quality with click-through on suggestions' sharpened UX debates measurably weekly. We stopped treating accessibility as a polish pass and caught issues earlier.
We should have invested in automated restore drills before the ransomware tabletop exercise exposed gaps. The design that considered assistive tech early avoided costly retrofitting later. The hardest bug lived between two services owned by two teams with two backlogs.
The migration that chunked batches avoided locking tables longer than maintenance windows. We stopped shipping 'temporary' SEO landing pages that contradicted the circles-first philosophy — brand confusion hurts conversion. We should have load-tested the auth path before Black Friday, not after.
We stopped treating 'innovation budget' as a blank cheque without expected learning milestones quarterly. The flaky integration lived between vendors; blameless vendor calls helped. The quiet win was documenting which Redis cluster is cache vs session vs rate limit — three names, three owners.
We finally aligned on a single severity rubric and stopped arguing adjectives. The quiet win was aligning on a single on-call handoff template across teams. We stopped shipping 'temporary' admin impersonation without audit logs — compliance nightmare.
We stopped confusing seniority with willingness to touch legacy code. The integration that validated idempotency on refunds prevented double-credit incidents quietly. The flaky deployment that ignored database migration order taught us to enforce ordering in CI.
Join the conversation.
Log in to reply