Three teams watch the same failure and count it three ways. The support team counts a ticket. The QA team counts a leaked defect — a bug that escaped testing and reached production. The customer success manager counts a churn-risk moment: a customer who just lost trust in the platform.
They’re the same event. And the fact that organisations measure it three times, in three departments, usually means nobody owns preventing it. At Cellulant, I made owning it my job description.
Owning the loop
My role there is, on paper, QA. In practice it’s the feedback loop between customers, support, and engineering: a user-reported issue comes in, I diagnose it directly across backend services, APIs, mobile apps and pipelines, and then — the part that matters — convert it into regression tests and shipped product fixes rather than one-off patches. Every customer-reported bug should die twice: once in the fix, once in a test that guarantees it stays dead.
The results of running that loop with discipline, from my CV: recurring production defects down 20% within six months, and automated test coverage across core services up 10% — bought by promoting clean-code practices and test-pyramid thinking, not by adding headcount.
Making quality visible
The other lesson from Cellulant: prevention work dies in spreadsheets. I led the design and build of a stakeholder-facing QA dashboard that aggregates real-time data from ClickUp and TestRail — the tools where the work already lives — and translates technical signals into terms the business can act on. Weekly reporting effort dropped, but the bigger change was what happened in meetings: quality trends became a standing slide instead of an anecdote.
Why CSMs should steal this
If you’re on the customer success side, you don’t need to run the pipeline — you need to read it. Ask your engineering partners: which root-cause categories are trending? Which accounts are generating repeat incidents? A rising repeat-incident count on a specific account is a leading indicator of churn, and it arrives months before the renewal conversation makes it obvious.
Support teams count tickets. QA counts bugs. CSMs count churn risk. The org that connects the three numbers prevents all of them.