Çelik Kasacı: a rebuild that made the numbers honest

Çelik Kasacı had the problem most growing stores have and few name out loud: the business was working, but nobody trusted the numbers it produced. Decisions were being made on data that quietly disagreed with itself, which meant they were really being made on instinct with a dashboard for cover.

The engagement was two jobs that turned out to be one: rebuild the store so it was structurally sound, and fix the tracking so the numbers told the truth. You cannot do the second on top of the first if the first is broken, which is why so many tracking fixes do not hold.

What was actually wrong

The reported revenue did not match what was landing in the bank. Ad platforms and analytics told three different stories. The team had stopped believing any single source, so every meeting started by relitigating whose number was right before anyone could decide anything.

Underneath, the causes were ordinary: events firing from more than one source, purchases counted more than once, and a store structure that had grown by accretion rather than design. None of it was exotic. All of it compounded.

The store was not failing. It was succeeding on data nobody could trust, which is a quieter and more expensive kind of problem.

What I did

First, the structure. I rebuilt the store so the foundation was clean: a sensible layout, a checkout that behaved predictably, and a data layer that emitted one clear signal per action instead of several muddy ones.

Then the tracking. One source of truth per event, deduplicated with stable identifiers, so a single purchase was counted exactly once. The analytics, the ad platforms, and the actual sales were brought into agreement, not by forcing them, but by removing the duplication and the gaps that had pushed them apart.

What changed

Concretely: reported revenue started matching the bank. Every purchase now fires from one source and is counted exactly once, so the duplicate-driven inflation is gone. Analytics, the ad platforms, and the actual sales finally agree, which means the weekly review stopped opening with an argument about whose number was right and started opening with a decision.

That is the measurable shift. The team stopped budgeting ad spend against numbers they did not believe and started trusting their own reporting enough to act on it the same day. Clarity is not cosmetic here. It is the difference between a brand run on evidence and one run on educated guessing with extra steps, and it was the rebuild and the tracking fix together, not either alone, that made it hold.