How to Reduce Stockouts: A DTC Brand's Guide
Stockouts get told as a good-news story. We sold out. Demand outran supply. The product is flying.
Read the numbers and a different story shows up. A stockout is a planning problem wearing a revenue costume, and for most scaling DTC brands it is one of the two largest inventory costs in the business, the mirror image of the dead stock sitting three shelves away. The same failure produces both: capital in the wrong SKUs.
This guide covers what a stockout actually costs, what genuinely causes them, and the weekly system that reduces them without drowning the business in safety stock it cannot fund.
What a stockout actually costs
Count it properly before fixing it, because the honest number is what justifies the fix.
For each SKU, take the days it was out of stock, multiply by the average daily units it sells when in stock, and value the lost units at gross margin, not revenue. Revenue overstates the loss; margin is what you actually gave up. Then be honest about recovery. Some buyers wait and purchase on restock. Some substitute another of your SKUs. Some leave and buy elsewhere, and only that last group is a true loss, so the real figure is a share of the gross calculation. We covered the full method inside the inventory audit guide, where stockout cost sits alongside dead stock as the two numbers that reframe inventory from a storage question into a capital allocation question.
Two costs sit outside that arithmetic and are routinely missed. Paid traffic that lands on an out-of-stock product page is ad spend converted directly into nothing, and it keeps spending until someone pauses the campaign. And a broken size curve is a stealth stockout: the product page looks alive, but if the two sizes that carry most of the demand are gone, the SKU is effectively out of stock while reporting as in stock.
The totals are rarely small. In one recent audit of a multi-brand consumer group, the avoidable stockout premium alone ran to roughly €260,000 a year, one line among €2.7M+ of identified annual margin. To run the same arithmetic on your own numbers, the stockout cost calculator in our tools hub does it in a couple of minutes.
What actually causes stockouts
Almost every chronic stockout pattern traces to a small set of root causes, and demand spikes are the rarest of them.
The most common is the absence of a reorder trigger. Nobody is comparing each SKU's cover against its true lead time, so the order goes in when someone notices the shelf, which is weeks too late by definition. Any SKU whose cover is shorter than its lead time is already out of stock; the system just does not know it yet.
Second is understated lead time. The supplier quote says four weeks. The true journey, production plus transit plus customs plus receiving, is nine or ten. Plans built on the quoted number fail on schedule.
Third is stale run rates. Averaging demand over weeks that include out-of-stock periods understates true demand, which flatters the cover figure, which delays the reorder, which causes the next stockout. The metric quietly feeds the problem it is supposed to prevent.
Fourth is cash. The brand knows it should buy more of the winner and cannot fund the purchase, usually because the money is trapped in slow stock bought two seasons ago. This is the cause that makes stockouts and overstock one problem rather than two.
Fifth is structural: minimum order quantities that force big infrequent buys, restocks competing with new launches for the same production slots, and concentration risk when one bestseller carries the catalogue and any wobble in its supply becomes a company-level event.
The system that reduces them
Not a project. A weekly rhythm, seven parts, most of which take minutes once built.
Start with a cover-versus-lead-time sheet, per SKU, refreshed weekly. Units on hand and in transit, average weekly sales over recent in-stock weeks, weeks of cover, true lead time, and the gap between the two. Sorted by the gap, the top of the sheet is the reorder queue. This single artefact replaces the noticing-the-shelf method and is the highest-leverage hour in DTC operations.
Measure true lead times rather than quoting them. Take the last five purchase orders per supplier and count the days from order to sellable receipt. Use that number, not the contract number, and remeasure quarterly, because lanes drift.
Set the reorder trigger at true lead time plus a buffer. The buffer is sized to how volatile the SKU's demand is, and it belongs on your A-items, the small cluster doing most of the revenue. Blanket safety stock across the whole catalogue is how the cash cause above gets worse.
Build the forecast the demand deserves. For core replenishment lines, a simple moving average over in-stock weeks beats intuition. For seasonal and trend lines it needs more care, and the methods worth using are in our inventory forecasting guide.
Use freight modes deliberately. Sea carries the planned core on a calendar; air is the emergency valve, reserved for fast restocks of proven winners where the premium buys revenue that would otherwise be lost. That is air freight used on purpose, the opposite of air freight as the invoice reactive planning sends you. The economics of that split are in the air versus sea cost breakdown.
Negotiate supply-side flexibility. Split shipments so the first tranche lands early. Rolling purchase orders that reserve production slots before final quantities are fixed. Smaller MOQs on colour and size variants even at slightly worse unit cost, because the markdown on the excess usually costs more than the unit-price saving.
And protect the size curve. Track stock at the size level on A-items, and treat the core sizes running dry as the stockout it is, whatever the product page says.
The causes, mapped to their fixes
Root cause | The fix | The metric that catches it early |
|---|---|---|
No reorder trigger | Weekly cover vs lead time sheet, per SKU | Count of SKUs with cover below lead time |
Understated lead time | Measure last five POs, use actuals | Quoted vs actual lead time gap |
Stale run rates | Average over in-stock weeks only | Demand rate refreshed weekly |
Cash trapped in slow stock | Clear dead stock, fund the winners | Dead stock at cost vs open-to-buy |
MOQs and big infrequent buys | Split shipments, rolling POs | Units per inbound shipment trend |
Launches crowding restocks | Separate production slots for core | Restock POs delayed by launches |
Broken size curves | Size-level tracking on A-items | Core-size availability rate |
When a stockout is the right call
Honesty requires this section. Some catalogues are built to sell out. Limited drops, deliberate scarcity, hype-led launches, running out is the model, and fighting it with replenishment logic wrecks the thing that makes it work. The discipline there is separation: split the catalogue into the drop side, where sellouts are the strategy, and the core side, where availability is the strategy, and never let the drop culture set the rules for the core. The brands that struggle are the ones applying one logic to both.
Frequently asked questions
What is an acceptable stockout rate?
There is no universal figure, and a blended rate hides the truth anyway. Track days out of stock on your A-items specifically. Core products that carry the revenue should be measured against a near-zero standard; the long tail can run leaner deliberately.
How much safety stock should we hold?
Enough to cover demand variability across the lead time, on A-items, bounded by cash. In practice: the more a SKU's weekly sales swing and the longer its lead time, the bigger the buffer, and a stable core line on a reliable lane needs surprisingly little. Resist formula-first approaches that spread buffer across the whole catalogue.
Are stockouts worse than overstock?
They are the same failure wearing different costumes: capital allocated to the wrong SKUs. The stockout costs you margin today; the overstock costs you cash, storage and markdown tomorrow. Fixing the allocation fixes both, which is why treating them as separate problems produces such poor results.
Can software solve stockouts?
Tools help once the discipline exists, and are noise before it does. The cover-versus-lead-time comparison is a spreadsheet; if the business is not acting on that, a planning platform gives it prettier numbers to ignore. Fix the trigger and the lead times first, then let software scale the routine.
