The pilot
The point is to see whether the analysis is useful on your stock, in your season, with your Flash Speed export. Workflows stay as they are while we do that.
How it runs
We do not connect to purchasing, we do not write back into Flash Speed, and we do not place orders. The analysis follows predetermined rules and produces recommendations for review.
Twelve to twenty-four months of sales or usage history, plus inventory on hand where it is available. CSV or Excel is fine. A first sample of 100–250 active SKUs is enough to show the method.
Each item is classified: runner, seasonal, intermittent, project/bulk, or slow/dead. Seasonality, ordinary usage, and exceptional transactions are separated so a bulk buy does not rewrite the baseline.
You get a report you can walk through with the people who already know the shop. They accept, reject, or adjust every recommendation. Nothing changes unless they decide it should.
What we need to start
A short data review comes first. That conversation confirms what Flash Speed can provide and whether the pilot is practical for your shop.
What does not happen
During the pilot, buying stays where it is. We are testing whether classification and seasonally adjusted usage produce recommendations a manager would actually use — not installing software into the counter or the warehouse.
If the report is useful, you decide the next step. If it is not, you have spent a sample of data and a review meeting, not a system project.
Send a note. We will tell you what we would need from Flash Speed and what a first sample should include.