What moved profit, and what to do about it
Cockpit always compares the trailing thirty days against the thirty before it — a fixed window, independent of whatever date range you've set elsewhere — breaks the change into the drivers that caused it, and pairs the waterfall with a short list of rule-triggered actions — not a suggestion engine, just the thresholds your own numbers crossed.
The biggest mover is routinely not the story
Order volume, average order value, product cost and payment fees each get their own bar against the center line, so a swing in one driver doesn't hide a swing in another — and whether the gains covered the losses is the question no single bar can answer alone.
- ✓A profit waterfall, thirty days against thirty days — the same window compared against the one immediately before it, biggest mover first.
- ✓Every ecommerce contribution margin driver isolated — order volume, average order value, product cost, payment fees, packaging, shipping net, ad spend, fixed overhead and other costs, each its own bar.
- ✓Gains against losses, stated plainly — total gained and total lost are called out separately, since a big green bar can still lose against a bigger red one.
- ✓Rules over your own numbers — each action lists the threshold it crossed, not a generic tip; a shop with nothing wrong gets nothing here.
- ✓Five actions, ranked — from unmapped purchase costs to concentration risk to unclassified statuses, each with a direct link to the report that explains it.
Nine drivers, ranked by size of the move, from order volume to average order value
Order volume added 6808.77 EUR, product cost took 4289.60, average order value took another 2192.52, payment fees a small 46.17 — with packaging, shipping net, ad spend, fixed overhead and other costs flat at zero this window. Read together: 6808.77 gained against 6528.29 lost, a net that only exists because the gain slightly outran the losses.
- ✓Order volume — the largest single mover this window, +6808.77 EUR.
- ✓Product cost — the largest drag, -4289.60 EUR.
- ✓Average order value — the second drag, -2192.52 EUR.
- ✓Payment fees — a small but real cost, -46.17 EUR.
- ✓Everything else flat — packaging, shipping net, ad spend, fixed overhead and other costs each moved 0.00 EUR this window.
Five rules, run over your own numbers
Each recommendation names the threshold it crossed and links straight to the report that explains it — from an unmapped cost basis, to a profit driver that went negative, to a customer base concentrated in too few accounts. A shop with nothing wrong gets nothing here.
- ✓Data health first — if purchase costs aren't mapped, no margin figure downstream can be trusted, so this rule fires before any other.
- ✓Negative drivers surfaced — any driver from the waterfall above that went negative gets its own action, linking to Economics.
- ✓Concentration risk named — revenue share from the top customers, checked against a fixed threshold, links to Customers.
- ✓Segment priority — the segment with the highest expected value is called out by name, linking to Segments.
- ✓Unmapped statuses and payment methods — anything still unclassified is flagged, linking to Economics settings.
The waterfall only shows what it can prove
Two guardrails run behind every render. First, a reconciliation check: the drivers you see must sum to the exact profit change the warehouse recorded — if they don't, Cockpit shows a mismatch warning instead of a number that looks right but isn't. Second, a data floor: with fewer than four daily data points or ten realized orders in either thirty-day window, Cockpit won't render drivers or actions at all — it says there isn't enough data yet, rather than showing a waterfall built on too few orders to mean anything. A cost that's genuinely missing (not mapped, not measured) is left out of the chart entirely, never drawn in as a silent zero.
- ✓Reconciliation, not just arithmetic — the displayed drivers must sum to the real recorded profit change, or a mismatch warning replaces false confidence.
- ✓A data floor before any driver renders — under four daily observations or ten orders in either window, Cockpit says so instead of guessing.
- ✓Missing costs are never zero — a driver with no data is omitted from the chart, not drawn in as "nothing happened."
- ✓Hover for the exact numbers — the profit-change line shows precise values and the exact date windows being compared on hover.
Know what moved profit, and what to do next
A thirty-day waterfall of every driver, plus a short list of rule-triggered actions — each with the threshold it crossed.
Ready to see what moved your profit?
Free check · 7-day trial · no credit card