Every inventory engagement starts with a 30-day paid backtest. We replay 24–36 months of your demand and supply history and show you, decision by decision, what we would have ordered — and what it would have meant in dollars.
Most inventory pitches end with a slide that says “our customers achieve 20% inventory reduction” or “15% stockout reduction.” That's not proof. That's an average from someone else's data.
A backtest on your data, replayed week by week, produces something a CFO can actually audit: a per-decision ledger of dollars. On these specific 47 SKUs in week 34, we would have ordered 800 units; you ordered 200; you stocked out three weeks later and lost €12,000 in revenue. Multiply across 24 months and the number stops being a marketing claim and starts being a verifiable record.
If the math doesn't work on your history, it won't work going forward. Better to know in 30 days than in 12 months.
24–36 months of sales transactions at SKU-day or SKU-week granularity. The more, the better — long history makes seasonal and trend signal much easier to separate from noise.
24–36 months of purchase orders received: order date, requested delivery date, actual delivery date, quantity ordered vs. quantity received. Lead time and reliability live in this data.
Stock levels over time, ideally at month-end or week-end. Used to compute true cover, working capital, and what your past decisions actually were.
Currently outstanding orders, plus a basic SKU master: pack sizes, units of measure, supplier, lead time targets, any explicit Stock Availability targets you already track.
No ERP integration. No software install. CSV exports — the same ones your team already runs for monthly reporting.
Every SKU-week in the backtest window where you stocked out. For each event:
Every SKU sitting on more cover than it should. For each:
A line-by-line CSV: every order decision, our recommendation, the dollar delta — totalled. Open it in Excel. Sort by SKU. Audit any line.
Three numbers compared against the success bar agreed at the start: Stock Availability uplift, inventory reduction, waste reduction. Plus a clear go/no-go recommendation on whether a recurring engagement makes sense.
Before the backtest starts, we agree in writing on the bar that defines success. The standard formulation is at least two of three thresholds:
If we hit at least two of the three on your historical data, we move to a recurring engagement.
If we don't, you keep the analysis and the reports. We walk away. No follow-up, no upsell.
The bar can be tightened or relaxed depending on your situation — what matters is that it's defined before the work starts, not after.
Send a message describing your situation — SKU count, ERP, current pain points. We'll get back to you to scope the data and the success bar. No pitch deck.
Contact Us