Blog · 20 September 2026
Out-of-stocks in click and collect and home delivery: why do they cost more than in store?
In shortIn store, the shopper decides alone in front of the empty shelf: substitute, delay or give up. In click and collect and delivery, the picker discovers the out-of-stock and chooses the replacement, and the shopper finds out at collection or at the door. The missing product is then a broken promise, not a mere gap on the shelf: refused substitute, refund, call to customer service, and the next basket placed elsewhere. The same store causes (replenishment, wrong book stock, badly prepared promotion) produce online a markedly higher cost per out-of-stock, because it includes handling the order.
A different out-of-stock because it was promised
On the shelf, the shopper sees the gap and chooses. The study by Gruen, Corsten and Bharadwaj measures the reaction: 26% switch brand, 19% take another size, 15% delay, 31% buy elsewhere, 9% give up. The store loses part of the sale, but the shopper kept control.
In click and collect or delivery, the order was confirmed and paid against book stock. The out-of-stock is found during picking, often by an employee who is not the aisle’s, and the choice of substitute falls to them. The shopper sees the result at the boot of the car or at the door. Three differences follow:
- the out-of-stock is a broken promise: the order showed the product as available;
- the substitution is imposed: a badly chosen replacement is refused, refunded, and costs a customer service contact;
- the out-of-stock is counted: every incomplete order leaves a trace, unlike the shelf gap nobody writes down. In the UK, the Retail Economics and DHL audit finds an out-of-stock item in one grocery trip in five; online, every one of those becomes a line on the order.
The real cost of an online out-of-stock
For a shelf out-of-stock, the cost is the share of the sale really lost, measured by the reactions above (how to measure sales lost to out-of-stocks). For an online out-of-stock, add:
| Item | What it costs |
|---|---|
| Picking time | looking for the product, choosing and validating the substitute |
| Order handling | line to remove, refund, credit, call or message |
| Refused substitute | product picked for nothing, sometimes a fresh item to throw away |
| Incomplete delivery or collection | the shopper goes elsewhere for the missing product, and often for the rest |
The complete-order rate is therefore the indicator that counts online, more than the out-of-stock rate per SKU. A 40-line order with one out-of-stock is an incomplete order for the shopper, even at 97.5% of lines served.
Where online out-of-stocks come from
The causes are the store’s, and Corsten and Gruen locate 72% of them in store: local order too tight, late replenishment, phantom stock. But online amplifies them:
- book stock is taken as truth: the site shows as available everything the system believes in stock; every phantom stock becomes a broken promise (phantom stock: what to do when the book stock says the product is there);
- shelf picking shares stock with in-store shoppers: an order picked at 4 pm against stock counted in the morning finds the shelf emptied by Saturday (Saturday and pre-holiday out-of-stocks);
- the promotion is ordered online without a cap: the leaflet attracts more orders than the promotional stock can serve (promotions and out-of-stocks).
What reduces online out-of-stocks
- Make book stock reliable on online SKUs: cycle counting targets first the fifty most ordered references online (cycle counting: twenty minutes a day).
- Pick before the peak: Saturday orders picked before 11 am find the shelf full.
- Define substitutes in advance: for each frequent product, a substitute validated by the aisle (same brand, other size first), rather than a choice improvised by the picker.
- Cap the promotion online: a maximum quantity per order on leaflet products.
- Feed online out-of-stocks back to the aisle: every unserved line is information for the department manager, the same day, not at the end of the week.
A worked example (hypothetical)
Take a fictional case. Supermarket with click and collect, 300 orders a day, 35 lines per order on average. Lines served: 97%, which looks good. But at one out-of-stock per order on average, 63% of orders are incomplete. After a month of cycle counting on the 50 most ordered SKUs and morning picking of Saturday orders, lines served rise to 98.5% and the share of complete orders goes from 37% to 59%. Customer service receives half as many complaints. The figures are illustrative.
What shopper reporting changes
In-store shoppers who report a missing product with ShelfAlert warn the online operation before picking: a product reported absent from the shelf at 3 pm is a product to remove from the site or substitute deliberately, not to discover at the shopper’s car boot. The aisle and the online team get the same information at the same time. 14-day trial, no payment card: see the plan for store managers.
Sources
- Gruen, Corsten, Bharadwaj, “Retail Out-of-Stocks: A Worldwide Examination”, GMA, 2002: full study (PDF).
- Corsten, Gruen, “On Shelf Availability”, 2004: summary.
- Retail Economics and DHL, The availability effect report 2026.
- DeHoratius, Raman, “Inventory Record Inaccuracy: An Empirical Analysis”, Management Science, 2008: abstract.