Why do my forecast numbers look wrong?

Work down this list in order for the variant you are looking at. The checks are ordered cheapest first, and most numbers that look wrong are explained in the first three.

1. When did the forecast last run?

The page shows the last run, not live data. Anything you changed since then - an order placed, stock received, a setting adjusted - is not in the numbers yet.

Press Run forecast, wait for it to finish, and look again. If the app is showing Your forecast is out of date, this is certainly the cause.

See The forecast is a snapshot.

2. Is the variant in scope?

A variant that is archived or draft in Shopify, or excluded by your forecast scope, behaves differently from one that is simply not selling.

Check Forecast & replenishment → Manage settings → Scope  for what is included.

3. Which location are you looking at?

Per-location numbers and All locations are different views of different data. A reorder point set for one location does not appear against another, and a suggestion that looks absent at All locations may be sitting under a specific one.

Confirm the location selector matches the location you are asking about.

4. Does the lookback period cover a representative stretch?

Sales velocity is measured over your lookback period. A 30-day window ending in a quiet month makes a seasonal product look dead; a window covering a single promotion makes it look like a runaway.

Check the lookback on the variant, then on your store default - the variant may carry its own.

5. Is a demand adjustment applied?

A demand adjustment scales velocity up or down by a percentage, and it stays applied until removed. An adjustment set for a promotion months ago is still inflating every number built on it.

Check both the store default and the variant override.

6. Was the variant out of stock during the period?

Velocity counts only the days a variant was in stock. A product that sold 30 units in the 3 days it was available before selling out has a velocity of 10 a day, not 1.

That is deliberate - it stops a stockout from suppressing future orders - but it does make a briefly-available product look far faster than a merchant expects.

If the variant is set to Continue selling when out of stock in Shopify, no days are excluded and the whole period counts, which produces the opposite surprise.

7. Is incoming stock netting off the suggestion?

Stock already on the way can reduce or remove a suggested quantity. If a suggestion is smaller than you expect, check what Stockie believes is incoming - see Incoming inventory.

8. Is there a min/max level on this variant?

A variant with min/max levels set uses them instead of the forecast, at the location they were set for. If a number ignores your reordering settings entirely, this is usually why.

Check Forecast & replenishment → Manage settings → Min/max levels , at the right location.

9. Do MOQs or pack sizes round it up?

A supplier minimum order quantity or a pack size rounds a suggestion upwards. A calculated 14 units from a supplier shipping in boxes of 12 becomes 24.

Check Forecast & replenishment → Manage settings → Supplier MOQs .

10. Is the demand actually recorded as sales?

Some movements never reach the velocity calculation:

• Manual inventory changes in Shopify move stock but record no sale.
• Unpaid orders are excluded - anything Shopify shows as pending, voided or expired.
• Wholesale draft orders invoiced but not paid sit as pending, so trade volume can be missing from a forecast that looks right for retail. Counting them is possible but not switched on by default - contact support.
• Cancelled orders are excluded, and returns reduce the quantity counted.

A variant whose sales come mostly through one of these routes will forecast low, correctly, from what it can see.

11. Is there any history to forecast from?

A new variant with no sales cannot have a velocity, so it has no reorder point and no suggestion. That is an absence of data rather than an error. See Forecasting new products.

Still wrong?

If you have worked through the list and still disagree with a number, contact support with:

• The SKU and the location
• What the number says, and what you expected
• Why you expected it - the sales you are comparing against, and over what period

That is enough for us to trace the variant through the same calculation and tell you which input produced the difference.