The most expensive sentence in e-commerce is "we're out of stock." It never arrives during a slow week. It arrives the moment your ads finally convert, your influencer posts, your email flow starts printing. The demand you paid for shows up, and your shelf is empty.
A stockout is not a logistics hiccup. You keep paying for clicks that land on a dead product page, your ad account's learning resets while you pause campaigns, and the customers who wanted to buy right now mostly do not come back. Stores that scale past the first phase treat inventory as a system with math attached, not as a pile of boxes they count when something feels low.
The Anatomy of Inventory Forecasting
Forecasting sounds like a finance department job. It is three numbers multiplied together. You need your average daily sales, your supplier's real lead time, and a buffer for the days when both of those numbers lie to you.
Know Your Real Lead Time
Lead time is not what your supplier quotes. It is the number of days between the purchase order and sellable stock on your shelf: production, freight, customs, receiving, and the day your 3PL takes to check it in. Track the promised date versus the actual date on every order. Most founders discover their real lead time is 30 to 50 percent longer than the quoted one.
Set Safety Stock Like an Operator
Safety stock is your insurance against variance, and two weeks of average demand is a sane starting point. Raise it for products with volatile sales or long overseas lead times, lower it for stable sellers with domestic suppliers. The cost of holding two extra weeks of a $4 item is trivial next to the cost of pausing campaigns that spend $200 a day.
Calculate the Reorder Point
Once you have daily sales and real lead time, the reorder point stops being a guess:
average_daily_sales = units_sold_last_90_days / 90
lead_time_demand = average_daily_sales * real_lead_time_days
safety_stock = average_daily_sales * 14
reorder_point = lead_time_demand + safety_stock
# Example: 12 units/day, 45-day real lead time
# reorder_point = (12 * 45) + (12 * 14) = 708 units
# When stock hits 708, you order. Not when it looks low.Recalculate monthly, and always recalculate before you raise ad spend. A scaling campaign changes your daily sales number, which quietly moves your reorder point up. This is the exact moment, covered in our Meta scaling playbook, when most stores walk into a stockout.
One warning in the other direction: forecasting cuts both ways. Ordering three months of stock on a product whose ads are still unproven is how cash dies on a shelf. Reorder points protect proven sellers. For anything still in testing, order small, accept the higher unit cost, and let the demand data earn the big purchase order.
Dealing with Suppliers and Quality Control
Suppliers are partners until the first delay, and then you find out what the relationship is actually made of. Two rules protect you. First, never let one factory hold your entire revenue: qualify a backup supplier before you need one, with samples approved and pricing agreed, so switching is an email instead of a panic. Second, put quality in writing. Order samples from every production run above a few hundred units, and inspect them against the original golden sample, not against your memory of it.
On payment terms, the standard 30 percent deposit and 70 percent after inspection exists for a reason. Paying everything upfront removes your only bargaining power when a container arrives with the wrong stitching. And when delays happen, because they will, renegotiate the next order's pricing or payment terms rather than accepting an apology. A two-week delay that costs you a product launch is a real number; treat it like one.
Watch the landed cost, not the unit price. Freight spikes, currency moves, and customs reclassifications can add 15 percent to a shipment without the factory changing anything. The fee stack from our cross-border costs breakdown applies to what you pay suppliers just as much as to what customers pay you.
Managing Backorders and Pre-Orders
A stockout handled well is a cash flow event, not a crisis. If your supplier has confirmed a ship date, switch the product page to pre-order with the honest date in the first line of the description, and offer something small for the wait, typically 10 percent off or free shipping. A clear date converts surprisingly well. Vague language like "back soon" converts nobody and generates support tickets.
If you cannot promise a date, do not take the money. Swap the buy button for a notify-me email capture and treat that list as your warmest launch audience when stock lands. Either way, message existing pre-order customers before they message you: one honest update at the halfway point of the wait cuts refund requests dramatically, and the communication templates from our support playbook handle the tone.
The One-Week Inventory Checklist
- Day 1: pull units sold per product for the last 90 days and compute average daily sales for your top ten SKUs
- Day 2: compare promised versus actual delivery dates on your last five purchase orders to get your real lead time
- Day 3: compute reorder points with the formula above and set low-stock alerts in Shopify at those numbers
- Day 4: email a second supplier for quotes and samples on your two best sellers, even if you never switch
- Day 5: write your pre-order page template and your two stockout emails before you need them
- Day 6: add a recurring monthly calendar event to recalculate reorder points and review ad spend plans against stock
- Day 7: count your best seller physically. The number in Shopify and the number on the shelf disagree more often than you think
Boring Operations Win
Reliable inventory is not glamorous, and that is the point. The stores that survive are not the ones with the cleverest ads; they are the ones that can actually ship what the ads sell. Three numbers, one backup supplier, honest dates, and a monthly recalculation. Do the boring work and stockouts become something that happens to your competitors.
Frequently Asked Questions
How much safety stock should a small store keep?
Start with two weeks of average demand per SKU. Raise it for products with volatile sales or lead times over 30 days, and lower it for stable sellers with domestic suppliers. Recalculate monthly, because your ad spend changes the number.
Should I take pre-orders or just mark the product sold out?
Take pre-orders only when your supplier has confirmed a ship date and you can print that date on the product page. If you cannot promise a date, use a notify-me email capture instead; taking money without a date creates refunds and disputes.
What do I do with a supplier who keeps missing deadlines?
Track promised versus actual dates on every order. If lead time slips more than 20 percent twice in a row, start moving volume to your backup supplier, but only after the backup's samples pass your quality check. Never cancel the primary before the backup is proven.



