Profitability
audit.
Four systems, four different answers to the same question: what did we sell, and what was left of it? The audit reconciles them and names the places where money is leaving quietly.
The problem it solves
Most retailers know their revenue to the zloty. Far fewer know what they actually earn once ad spend, returns, marketplace fees, shipping and FX have been taken out — and almost nobody knows it broken down by channel, by market and by SKU at the same time.
The gap is not laziness. It is that every system measures something slightly different. The advertising panel counts attributed sales in its own window. Analytics counts sessions and applies its own attribution. The order system counts orders, including the ones that were later refunded. The bank counts what actually arrived, weeks later, net of fees. Put four honest systems side by side and you get four different numbers.
What happens
- Pull the raw data from each source rather than the summaries each panel offers.
- Reconcile them against one another over a common period, order by order where possible.
- Rebuild contribution margin per channel, market and SKU with real costs attached.
- Trace each discrepancy to its cause — attribution window, refund handling, fee timing, currency, a broken integration.
What you get
- A discrepancy report: where the numbers disagree, by how much, and why.
- A margin view by channel, market and SKU for the audited period.
- A ranked list of the places where the business is losing money or acting on numbers that are not true.
- A short recommendation on what to fix first, and what it would take.
Who this is for
Businesses selling on more than one market and through more than one channel, without an in-house data team. If you have a data engineer and a warehouse already, you do not need this — you need someone to ask better questions of it.
Start here
The audit is deliberately small and time-boxed. It ends with a document, not a dependency — and it is the honest way to find out whether the deeper work is worth doing at all.