A dashboard can show a clean, confident-looking number even when the data behind it is wrong. A percentage does not announce that it was built from a stale stock count, a double-counted batch, or a location feed that never updated. The number simply displays, and a manager who has not learned to check it will trust it exactly as much as a number built from correct data.
This lesson mainly teaches the information flow behind a domestic stock-cover KPI: where each input comes from, how a data error can enter unnoticed, and how to confirm a number is trustworthy before a reorder decision is made on it. The case uses Cedar & Co., a domestic brand that manages its own stock from its own forecast, ordering from Delta Garments on a made-to-stock basis.
You need: the KPI-definition rule from Lesson 4.1 and the leading/lagging distinction from Lesson 4.2. You also need a basic understanding of a stock-cover calculation (stock on hand divided by average sell-through).
By the end of this lesson you can:
Not covered here: how to design a stock-cover formula itself; this lesson assumes the formula is correct and focuses on whether its inputs are correct.
| Term | Meaning |
|---|---|
| Data input | One of the individual figures that a dashboard calculation combines to produce its displayed number. |
| Stock cover | The number of weeks current stock would last at the current average sell-through rate. |
| Sell-through rate | The average quantity sold per period, used to judge how fast stock is being used up. |
| Stale data | Information that was once correct but has not been updated to reflect a more recent change. |
| Double count | An error where the same physical stock is included twice in a total, overstating it. |
| Source check | Comparing a dashboard number against the original record it was supposed to be built from. |
A calculated KPI, such as stock cover, is never more accurate than the data feeding it. If one input is wrong, current, and confidently displayed, the calculated result will also be wrong, current-looking, and confidently displayed. The screen gives no visual difference between a number built from good data and a number built from bad data; only checking the inputs against an independent source can tell them apart.
| What can differ | The shapes you will meet | Why it matters to you |
|---|---|---|
| Domestic, made-to-stock replenishment | Stock cover can be overstated when goods still in transit are counted as if already in the warehouse. | A reorder that should have been triggered gets delayed until an actual stock-out happens. |
| Export, made-to-order production floor | A cumulative-sewing percentage can be overstated when partly completed pieces are counted as fully sewn. | Overtime that should have been approved is not, and the port cut-off is missed instead. |
| Buyer or brand side | A fabric-receipt confirmation can be stale when it reflects an email from several days ago rather than the current warehouse record. | The buyer's leading KPI loses its whole value if the underlying confirmation is out of date. |
| Mixed export and domestic portfolio | A combined report can double-count a batch that appears in both an export shipment record and a separate domestic transfer record. | A manager comparing orders side by side may misjudge which one actually needs urgent attention. |
| Different product families | Hosiery stock counted by the dozen pairs can be entered inconsistently as individual pairs; bedding sets can be counted by set or by individual piece without a clear rule. | A unit-of-measure mismatch produces a wrong total even when every individual count was recorded correctly. |
Cedar & Co. is a domestic brand with 12 shops that manages its own stock from its own forecast, buying from Delta Garments on 60-day open account. Cedar & Co. has reordered a lot sized at 15,000 pieces, reusing Order A's own quantity, at Order A's own equivalent component cost of USD 2.45 per piece (its material, trims and sewing components added together). The dashboard shows a stock-cover KPI: weeks of stock remaining at the current average sell-through.
The dashboard currently displays 9 weeks of stock cover, calculated from a stock-on-hand figure of 9,000 pieces at an illustrative average sell-through of 1,000 pieces per week. Omar, the owner, is deciding whether to delay the next reorder because the number looks comfortable. Before trusting it, the merchandiser checks the stock-on-hand input against the warehouse's own physical count and the shipment record from Delta Garments, because 2,000 of those 9,000 pieces are still recorded as "in transit" in the logistics system, not yet received into the shop network's own stock.
The worked numberThe dashboard's displayed figure counts the in-transit pieces as if they were already available stock:
Once the source check removes the 2,000 pieces still in transit, the true available stock is only 7,000 pieces:
The difference between 9 weeks and 7 weeks is exactly the two weeks a reorder decision needs to be made in time. If Cedar & Co.'s usual reorder lead time from Delta Garments is close to 7 weeks, trusting the uncorrected 9-week figure would mean placing the reorder too late, risking an actual stock-out before the new lot arrives.
A common mistake is assuming that a dashboard showing "9.0 weeks" rather than "about 9 weeks" must be accurate, because the precision looks technical and confident. Precision describes how the number is displayed, not whether its inputs are correct; a wrong input carried through a precise formula still produces a wrong, precisely displayed answer. The cost of this mistake is a decision, such as delaying a reorder, made on a number that felt more trustworthy than it actually was.
The better habit is to treat every dashboard figure used for a real decision as unconfirmed until its inputs have been checked at least once, and to recheck periodically rather than assuming a past check still holds.
A number on a screen looks the same whether it was built from correct data or from a mistake. Before you trust a number enough to make a decision with it, find out what it is actually made of: which figures were added, subtracted, or divided to produce it. Then check at least the figure most likely to be wrong against a different, independent record. A calculation is only as good as what goes into it, and a screen will never tell you on its own which numbers to trust.
Cedar & Co.'s dashboard shows 9 weeks of stock cover on the current 15,000-piece lot, based on a stock-on-hand figure of 9,000 pieces and an average sell-through of 1,000 pieces per week. Omar is about to approve delaying the next reorder from Delta Garments by three weeks, since 9 weeks looks comfortable. The merchandiser has just discovered that 2,000 of the 9,000 pieces are still recorded as in transit, not yet received.
Should Omar approve the three-week delay based on the dashboard's displayed 9-week figure? Decide before opening the answer.
Omar should not approve the delay until the stock-cover figure is recalculated with the in-transit pieces removed, because the corrected figure changes the decision.
1. The displayed 9,000-piece stock-on-hand figure includes 2,000 pieces still in transit, which are not available to sell from the shop network yet. 2. The corrected stock on hand is 7,000 pieces, giving 7 weeks of stock cover (7,000 ÷ 1,000), not the 9 weeks shown on screen. 3. A three-week delay on top of a corrected 7-week cover would leave only 4 weeks of buffer, not the 6 weeks Omar believed he was approving based on the uncorrected number. 4. The reorder decision should be based on the corrected 7-week figure, and Omar should ask whether a 3-week delay still leaves enough buffer once Delta Garments' own reorder lead time is compared against it.
The lesson is not that Omar should never delay a reorder; it is that the decision must use the corrected number, not the number that happened to display first.
Cedar & Co.'s reordered lot is priced at USD 2.45 per piece delivered cost, on a 15,000-piece lot:
The 2,000-piece data error is only 22% of the displayed 9,000-piece stock-on-hand figure, but it changes the stock-cover result from 9 weeks to 7 weeks, a difference large enough to change a reorder-timing decision. If Cedar & Co. delays reordering based on the uncorrected 9-week figure and then experiences an actual stock-out before the next lot arrives, the cost is not the USD 36,750 of the lot itself; it is the lost sales during every week the shops have no stock to sell, plus any customers who buy from a competitor instead and do not come back. A stock-out is far more expensive to a domestic brand than the small effort of checking one dashboard input before trusting it.
Use the T17-C04 KPI and decision-data workbook. For each dashboard KPI you rely on, list every input it is calculated from, the source you would check each input against, and the date it was last checked. Start with Cedar & Co.'s stock-cover KPI and its in-transit correction, then repeat the exercise for one KPI already used in your own workplace.
[Download the KPI and decision-data workbook](../../templates/T17-C04-kpi-and-decision-data-workbook.xlsx)
1. Why can a dashboard display a wrong number with just as much confidence as a correct one?
2. On Cedar & Co.'s dashboard, why does removing the 2,000 in-transit pieces change the stock-cover result from 9 weeks to 7 weeks?
3. A sock brand's dashboard shows 6,000 pairs in stock and average sell-through of 800 pairs per week, but 1,200 pairs are recorded twice by two warehouse systems. What is the corrected stock cover?
| EN | AR | FR | TR | Meaning |
|---|---|---|---|---|
| Data input | مدخل بيانات | Donnée d'entrée | Veri girdisi | One of the individual figures that a dashboard calculation combines to produce its displayed number. |
| Stock cover | تغطية المخزون | Couverture de stock | Stok kapsama süresi | The number of weeks current stock would last at the current average sell-through rate. |
| Sell-through rate | معدل البيع الفعلي | Taux d'écoulement | Satış oranı | The average quantity sold per period, used to judge how fast stock is being used up. |
| Stale data | بيانات قديمة | Donnée obsolète | Bayat veri | Information that was once correct but has not been updated to reflect a more recent change. |
| Double count | احتساب مزدوج | Double comptage | Çift sayım | An error where the same physical stock is included twice in a total, overstating it. |
| Source check | التحقق من المصدر | Vérification de la source | Kaynak kontrolü | Comparing a dashboard number against the original record it was supposed to be built from. |
Next you will learn how to turn a management report's findings into named actions, owners and follow-up dates across a mixed portfolio of orders. Back to the course map