boundary
Should You Reorder on a Weak Forecast or Wait for Better Data?
When the history is thin, act anyway on the cheapest reversible order and write down the assumption you acted on, instead of waiting for a forecast that will not arrive. StockMosaic makes that trade visible: every proposal carries its assumptions, so you can see which lines rest on data and which rest on judgment. You can open these pages and sign in to the workspace today. That does not mean StockMosaic is connected to your e-commerce platform or point-of-sale system, and it does not mean the product has formally launched.
The trade-off this question is really about
The searcher is usually stuck between two fears: stocking out of something customers wanted, and parking cash in something that will not move. The honest answer is that no forecasting method rescues two weeks of data. What helps is separating the SKUs where the number is trustworthy from the ones where you are guessing, and sizing the guesses small enough that being wrong is cheap.
Forecasts depend on sales history, lead-time quality, and stated assumptions. Purchase orders remain proposals until a buyer approves them. The desk will not dress a two-week history up as a demand curve, and neither should anyone else.
What you need beyond the spreadsheet
Whatever history exists, the supplier's real recent lead times, and a rough sense of what being wrong costs in each direction for the SKUs in question. For most small retailers, a stockout on the bestseller costs more than a slow quarter on a marginal line, and that asymmetry should drive order sizes.
Step 1: Split the catalog into trusted and thin
Trusted means a stable selling rate over enough weeks to believe, and a lead time you have seen hold. Thin means everything else: new products, seasonal items entering their first season, SKUs whose sales spike once a month. Write both lists down; the split drives everything after.
Step 2: Let the ranking drive the trusted SKUs
For trusted lines, run the normal reorder-priority workflow and size orders off the ranked proposal. This is where the data earns its keep, so spend your judgment budget elsewhere.
Step 3: Size the thin-SKU orders to the downside
Order the smallest quantity that keeps the SKU alive past the next review point. If the item turns out to be a hit, reordering in two weeks costs you a little margin. Over-ordering a miss costs you the shelf space and the cash for months.
Step 4: Record the assumption beside every thin line
Each guessed order carries one sentence: what you assumed and why. 'Assumed two per week based on similar item's first month' is auditable later. A bare number is not, and unauditable guesses are how overstock happens twice.
Step 5: Revisit thin lines when real sales arrive
Set a review point at roughly one lead time after the goods land. Compare actual sales to the assumption, promote lines that held up to trusted, and shrink the thin set each cycle. The goal is not better guessing; it is needing to guess less.
Verification
You can say, for any order line, whether it came from trusted data or from a stated assumption, and quote the assumption. Thin lines are sized so that being wrong stings but does not damage. Each cycle, the share of trusted lines is going up; if it is not, the review step is being skipped.
Limits under thin data
Forecasts depend on sales history, lead-time quality, and stated assumptions. Purchase orders remain proposals until a buyer approves them. The desk does not discover demand patterns absent from the data you supplied, and an unstable supplier lead time is a fact to write down, not smooth over.
No output here removes the buyer's decision. It makes the decision easier to audit after the fact.
How StockMosaic supports the guesswork
StockMosaic holds the trusted-versus-thin split, attaches your stated assumptions to the thin lines, and flags which proposals rest on the weakest ground. Ask it to compare actuals against assumptions at each review point.
You set the risk appetite, you approve the orders, and you decide when a thin line has earned promotion.
FAQ
Questions this guide is for
Is two weeks of history enough to reorder at all?
For a fast mover with a stable supplier, yes: small first order, assumption written down. For a slow or seasonal SKU, no; wait for real sales rather than manufacturing a rate.
The supplier's lead time keeps changing. What do I use?
Use the recent worst case as the working lead time and flag it as an assumption. Forecasts depend on lead-time quality, so instability is an input, not an excuse.
Can StockMosaic auto-reorder for me?
No. Purchase orders remain proposals until a buyer approves them, and there is no supplier connection on this path.
Start in the workspace
Make the thin-data calls explicit
Sign in or create an account. You return to the StockMosaic conversation. Split your catalog, size the guesses small, and put the assumption on the record before the money moves.