AI use case · Is your data ready for demand forecasting?Demand forecasts learn from what you sold. Not from what you could have.
Demand planning tools promise fewer stock-outs and less excess stock. They learn from order history. If products were re-coded, stock-outs were never recorded and promotions live in a planner’s spreadsheet, the model mistakes lost sales for low demand and promotions for normal weeks.
Starts with a 30-minute call. Commits you to nothing.
What it needs
What demand forecasting needs from your data.
Each requirement, the systems it lives in, and what we check. The AI is the easy part; this list is the hard part. Each check feeds the Truth Score™, a score out of 100 for how far your data can be trusted.
Truth Score pillars that gate it mostCompletenessAccuracyConsistencyThe six pillars
Where it breaks
Every system looks fine. The gaps between them don’t.
Where demand forecasting breaks: not inside one system, but where two of them disagree about the same customer, deal or invoice.
-
Operations ↔ Finance
Items re-coded in the ERP, history split across old and new codes
What it does to the AIThe model sees a new product with no past.
-
Web ↔ Finance
Online orders that don’t match posted sales
What it does to the AITwo versions of demand, and the model picks one.
-
Spreadsheets ↔ Operations
Promotions planned in spreadsheets, never tagged on the order
What it does to the AIA promotion spike is learnt as normal demand.
Gaps to test for this use case, not client findings.
Before you sign
Questions to ask the vendor selling it.
Four questions that separate a demo from a deployment. Bring the answers, or the proposal, to the Truth Score Read and we test them against your records.
- Q1
How do you separate true demand from what was simply in stock?
If the answer is vague, the tool will under-order exactly the products you ran out of.
- Q2
How do you handle new and re-coded products?
Ask to see how history carries across a code change. Many tools treat it as a new item.
- Q3
What do you need from us about promotions?
If promotions aren’t on the order, someone will have to supply them, every time.
- Q4
Do you measure accuracy against a simple forecast?
Ask for the comparison with last year’s same period. A tool that can’t beat it isn’t worth the licence.
How a Truth Score Read tests it
Demand forecasting, tested against your own records.
Our engine matches the customers, people, deals and invoices across the exports from the systems behind the decision. Every red flag is checked by hand against sample records before sign-off, and the verdict comes in writing. Here is what that means for this use case.
- 01
Rebuild the history
Order and sales exports joined across channels and compared with posted sales in the ledger.
- 02
Stitch the product codes
Re-coded, superseded and duplicate items found, so you know how much of the history is really usable.
- 03
Look for the missing signals
Whether stock-outs, promotions and returns can be identified anywhere in the exports or spreadsheets.
- 04
Give the verdict
The level, item, category or site, at which a forecast can be trusted today.
Two systems are enough to start. We send the exact field list after the call.
How the Truth Score Read worksWhat you get back
See what comes back: a real sample memo.
One fictional B2B software company, Example Co: 72,837 records in 7 exports from its CRM, finance and HR systems, run through the same engine and the same checks as a paid Truth Score Read. Each system passes its own checks. Matched across systems, only 13% of deals tie to an invoice.
Questions
Demand forecasting, answered plainly.
Anything else, ask on the 30-minute call. It commits you to nothing.
Book the callWhat data does AI demand forecasting need?
A long, consistent order history at the level you plan, one product code across that history, and a record of stock-outs, promotions and returns. The last three are what separate real demand from what simply happened.
Why do stock-outs matter so much?
Because when a product is out of stock, sales fall to zero and the model reads that as no demand. Without a record of stock-outs, the forecast systematically under-orders your best sellers.
How do product code changes affect forecasting?
They cut the history in two. Unless old and new codes are linked, the model sees a new product with no past and an old one that stopped selling. Linking them is a data task, not a modelling one.
Does this apply to services businesses?
The same logic applies to capacity: bookings, appointments or billable hours instead of units. The history, the codes and the missing signals still decide what a forecast can do.
Is your data ready for demand forecasting? Find out before you sign.
Exports from two or more systems in. A signed verdict on this use case within 5 working days of complete exports. Starts with a 30-minute call. Commits you to nothing.
Starts with a 30-minute call. Commits you to nothing.