Case study · Supply chain · Live
Fill-rate root cause in 5 minutes, not a day
When a retailer order shipped short, working out why took an analyst four to eight hours of pulling and lining up reports. Now an agent does the digging, and the analyst reviews the findings and makes the call.
The problem
For a brand selling through retailers, a short-shipped order is expensive: missed sales, retailer scorecard hits and sometimes penalties. Every time it happened, someone had to answer the same question: why did we miss that order?
There are usually three suspects. Was the forecast off? Did a supplier deliver less than was ordered? Or was the stock on hand, but sitting in the wrong place? Answering meant pulling three or four reports and lining them up by hand, customer by customer and product by product.
It took an analyst four to eight hours, every time. By the time the answer was ready, the next order was already on its way.
What the agent does
- Works out fill rate by customer and product, as soon as the data lands.
- Finds the gaps and ranks them by how much they cost.
- Investigates each one and names the likely cause: a forecast miss, a supplier shortfall or stock in the wrong place.
- Suggests a fix for each gap, with the evidence behind it.
- Hands it to a person. The analyst reads the findings and decides what to do.
How the team stays in control
The agent doesn't change orders, forecasts or stock on its own. It does the digging, and a person makes the decision. Every run records what the agent looked at and how it reached each conclusion, so any finding can be checked.
The result
The review that used to take four to eight hours is now ready in about five minutes. The analyst's time goes into deciding what to do about each gap, not into building the spreadsheet that finds it.
“They were very quick to take us from the first discussion to design and into production. We saw impactful changes within weeks.”
What the analyst sees
A ranked list of gaps, causes and fixes
| Customer · product | Fill rate | Likely cause | What the agent found | Suggested fix |
|---|---|---|---|---|
| Retailer A · Product 1042 | 71% | Forecast miss | Sell-through ran well above forecast for three straight weeks. | Raise the next four weeks' forecast and pull the open PO forward. |
| Retailer B · Product 2217 | 84% | Supplier shortfall | Supplier delivered 1,200 of 2,000 units on the last PO. | Chase the 800-unit balance and flag the lead-time risk to planning. |
| Retailer A · Product 3380 | 88% | Stock in wrong place | Enough units on hand, but in the warehouse that doesn't serve this account. | Transfer stock before next week's order arrives. |
Customer, product and figures above are sample data to show the format. How agents catch supply chain problems →
Got a process like this one?
Tell us about it. In 30 minutes we'll walk through it with you and tell you plainly whether an agent can take it on. If it can't, we'll say so.