It's Friday evening. Your bestselling kurta just sold its last unit, your ads are still running, and your manufacturer's phone is off until Monday. If you're wondering how to avoid stockouts without the opposite problem — a warehouse full of cash you can't touch — this guide is the playbook. Nine strategies, each with the math behind it. No "just order more" advice.
Key takeaways
- A 2-week stockout on one bestseller can cost more than a quarter's software budget — count lost sales, wasted ad spend, and the recovery dip.
- Track daily velocity per SKU, and exclude stockout days from the average — or every stockout makes the next one more likely.
- Reorder points and safety stock turn "when should I reorder?" from a feeling into a number.
- Protect your A-grade SKUs first: roughly 20% of products drive 80% of revenue.
- Automate the reorder signal — humans forget to check spreadsheets; software doesn't.
What a stockout actually costs
Before the strategies, the stakes. Take a product priced at ₹1,200 that sells 5 units a day. It stocks out for two weeks — a very ordinary stockout for a brand with a 30-45 day production cycle. Here's the bill:
| Cost line | Calculation | Amount |
|---|---|---|
| Direct lost sales | 14 days × 5 units/day × ₹1,200 | ₹84,000 |
| Wasted ad spend driving traffic to a sold-out page | ~15% of period ad budget, e.g. ₹40,000 × 15% | ₹6,000 |
| Post-restock recovery dip (lost ranking, momentum) | ~30% lower sales for 7 days: 7 × 1.5 × ₹1,200 | ₹12,600 |
| Total for one SKU, one stockout | ₹1,02,600 |
Over a lakh, from a single variant going dark for a fortnight. And notice that the first row is the only one most founders count. The other rows are why stockouts hurt more than the revenue line shows: paid traffic keeps landing on a dead page, and the algorithmic momentum you bought has to be bought again after restock. Multiply the total by the three or four stockouts most growing brands hit per quarter and the case for fixing this makes itself. Now, the nine strategies.
1. Track daily sales velocity per SKU
You cannot avoid what you cannot see coming. The foundation of stockout prevention is knowing, for every SKU, how many units it sells per day — this week, not last quarter. Monthly reviews are too slow: a product doing 3/day in May and 6/day in June will be gone weeks before your next review.
Velocity turns stock levels into time. "42 units left" means nothing on its own. "42 units at 4.2/day = 10 days remaining" is a decision. That countdown — days remaining, per SKU — is the single most useful number in inventory management, and it's the core of how Honey Shelf's inventory engine works.
Two practical rules for computing it. Pick a window that matches the product: 30 days is a sensible default, 7-14 days when something is trending, 90 days for stable classics. And count units, not revenue — a price change should never look like a demand change.
2. Set reorder points, not gut feelings
A reorder point is the stock level at which you must reorder to receive goods before you run out. The formula is simple:
Reorder Point = (Daily Velocity × Lead Time in Days) + Safety Stock
Selling 4/day with a 14-day lead time and 20 units of buffer? Reorder at 76 units — not when the shelf "looks low." The point of the formula is that it removes the argument: when stock crosses the line, you order. We've written a full walkthrough with two worked examples in our guide to calculating reorder points.
3. Hold calculated safety stock — not vibes-based padding
Safety stock is the buffer that absorbs what averages can't: the viral reel, the supplier who ships a week late. Held on every SKU "just in case," it's how brands end up overstocked. Calculated per SKU, it's cheap insurance.
The workhorse version: (max daily sales × max lead time) − (avg daily sales × avg lead time). For most D2C brands that lands at 2-4 weeks of cover on bestsellers and near zero on C-grade slow movers. The full method, including the service-level approach, is in our safety stock formula guide.
4. Buffer for real lead times, not quoted ones
Your manufacturer says 21 days. Your last five POs arrived in 24, 22, 31, 25, and 28 days. Planning on 21 means you'll stock out two orders out of five. Use the actual distribution: plan on the average, buffer for the worst case you've seen in the last six months.
This requires keeping receipts — literally. Log promised date vs. actual date on every PO. It's tedious in a spreadsheet, which is why most brands don't do it, and why supplier lead-time tracking is built into Honey Shelf rather than left to memory.
5. Prioritize bestsellers with ABC analysis
You can't watch 200 SKUs equally, and you shouldn't. Rank products by revenue contribution: A-grade SKUs (typically the top ~20% of products driving ~80% of revenue) get daily monitoring, generous safety stock, and first claim on production capacity. B-grade gets standard reorder points. C-grade gets minimal buffer — a stockout on a product selling 0.2/day costs you almost nothing.
This is the "without overstocking" half of the equation. Brands that treat every SKU the same end up with too much buffer on slow movers and too little on the products that pay the bills.
Running the analysis takes an hour: export 90 days of sales by SKU, sort by revenue, cumulative-sum down the column. Draw the line at 80% for A-grade and 95% for B; everything below is C. Redo it quarterly — last festive season's hero is often this quarter's C-grade, and the grades move more than most founders expect.
6. Score your suppliers on on-time delivery
Lead-time buffers (strategy 4) treat the symptom. Supplier scorecards treat the cause. Track each supplier's on-time delivery rate (OTD) as a rolling percentage. A supplier at 94% OTD needs a small buffer; one who has drifted to 71% needs either a hard conversation or a replacement — because no formula can cheaply protect you from a supplier who's late one time in three.
Review scorecards quarterly. Falling OTD is a leading indicator: it shows up in your data months before it shows up as your worst stockout of the year. And share the numbers with the supplier — most partners tighten up quickly once they know late deliveries are being measured rather than merely grumbled about.
7. Correct your demand math for stockout days
Here's the trap that keeps brands in a stockout loop, and the reason "how to avoid stockouts" is partly a math question: a stockout poisons the very average you'll use to plan the next order.
Worked example. Your kurta sells 4 units a day when it's available. Last month it was out of stock for 10 of 30 days, so you sold 80 units (20 days × 4/day).
- Calendar average: 80 units ÷ 30 days = 2.7/day
- True velocity: 80 units ÷ 20 in-stock days = 4.0/day
Plan with 2.7/day and every number downstream — reorder point, order quantity, safety stock — is 33% too low. You'll stock out again, which drags the average down further, which makes the next order even smaller. Zero sales on a stockout day isn't zero demand; it's missing data. Exclude those days from the average. Honey Shelf does this correction automatically on every SKU, which is why its countdowns stay honest even after a bad month.
8. Sync your store and your production floor
Many "stockouts" are bookkeeping failures: the goods exist, but Shopify doesn't know. Finished units sit at the factory for a week before someone updates the store; or the store oversells because a manual count was stale. Every handoff that relies on a human typing a number into Shopify is a future incident.
Close the loop both ways: sales flow from your store into your planning system daily, and finished goods flow back as live inventory the moment production completes. That's exactly what two-way Shopify sync does — sizes, variants, and all.
9. Automate the signal
Every strategy above works on paper and fails in a spreadsheet, for one reason: someone has to remember to look. The founder gets busy in a launch week, the ops lead goes on leave, and the reorder point quietly slides past unnoticed.
The fix is to make the system come to you. Velocity is recalculated daily, the days-remaining countdown crosses your threshold, and a draft production order appears with quantities pre-calculated — you approve it or you don't. In Honey Shelf's 6-stage pipeline, that signal is stage one, and it fires early enough for your supplier's lead time, not just early enough to panic.
Putting it together: a 30-day rollout
You don't need all nine on day one. A sequence that works:
- Week 1: compute stockout-corrected velocity for your top 20 SKUs (strategies 1 and 7). Pure math on data already sitting in Shopify.
- Week 2: set reorder points and safety stock on your A-grade products (2, 3, and 5).
- Week 3: start logging promised-vs-actual dates on every PO and open supplier scorecards (4 and 6).
- Week 4: connect store and production data, and put the reorder signal on autopilot (8 and 9).
Then hold the line. Discipline that depends on a human checking a spreadsheet every morning is discipline that eventually fails — automate it and spend the attention on product instead. The goal isn't more inventory. It's the right inventory, ordered on time, every time.