The reorder point formula is not wrong. The demand rate people feed it is.
Reorder point equals demand during the lead time plus safety stock. The mistake is computing "demand during the lead time" from a trailing twelve month average, because the lead time you are covering in October does not experience average demand. It experiences November demand. A reorder point built on the annual average will trigger the order at the right stock level for a Tuesday in April and roughly half the right level for the week before Cyber Monday.
The correction is mechanical: compute the demand rate that will actually occur inside the window the order covers, not the demand rate that occurred on average over the past year.
The U.S. Census Bureau's Quarterly Retail E-Commerce Sales release for the first quarter of 2026, published 18 May 2026, gives the shape of the season at the market level. On a not-adjusted basis, U.S. retail e-commerce sales rose from 299.7 billion dollars in the third quarter of 2025 to 365.2 billion in the fourth quarter, an increase of 21.8 percent. They then fell 17.2 percent to 302.3 billion in the first quarter of 2026.
That is the whole-market average across every category, including ones with almost no seasonality. An individual seller in toys, electronics, or gifting runs a far steeper curve. If your own Q4 index is not materially above 1.2, either you sell something genuinely aseasonal or you are not measuring it.
You need one number per month: actual units sold, last year, per SKU or per SKU family. Take a real twelve months.
| Month | Units | Index |
|---|---|---|
| Jan | 1,300 | 0.65 |
| Feb | 1,200 | 0.60 |
| Mar | 1,400 | 0.70 |
| Apr | 1,450 | 0.73 |
| May | 1,500 | 0.75 |
| Jun | 1,500 | 0.75 |
| Jul | 1,550 | 0.78 |
| Aug | 1,600 | 0.80 |
| Sep | 1,800 | 0.90 |
| Oct | 2,500 | 1.25 |
| Nov | 3,800 | 1.90 |
| Dec | 4,400 | 2.20 |
| Total | 24,000 | |
Average month is 2,000 units. The index is each month divided by that average. November runs at 1.90 and December at 2.20, which means the daily rate in those months is roughly two to three times the rate the annual average implies.
Daily rates matter more than monthly ones because lead times do not respect month boundaries. October is 80.6 units a day. November is 126.7. December is 141.9. The annual average is 65.8.
Assume a 45 day supplier lead time and a safety stock policy of ten days of cover.
Using the annual average of 65.8 units a day:
Reorder point equals 65.8 times 45, which is 2,961, plus 65.8 times 10, which is 658. Total 3,619 units.
Now compute it properly for an order placed on 15 October 2026, arriving 45 days later on 29 November. The reorder point has to cover demand from the moment of ordering until the goods arrive, which is 17 days of October and 29 days of November.
Safety stock should be expressed in days at the rate that will be running when you need it, not the annual rate. Twelve days at the November rate is 1,520 units.
Seasonal reorder point: 6,564 units.
The flat-average number, 3,619, is 2,945 units short. Across a window running at roughly 110 units a day, that shortfall is about 27 days of demand. A seller who reorders at 3,619 units places the order 27 days too late and hits zero in the last week of November.
Last year's units are the starting point, not the answer. If the business is running 22 percent above last year on a comparable basis, scale the seasonal rates before you use them.
Recomputing the same 15 October order:
Against the flat-average 3,619, that is 2.2 times higher. Every unit of that difference is a unit you either have or do not have on Black Friday.
Growth rate deserves scrutiny before you multiply by it. Growth that came from adding a channel does not repeat. Growth that came from a competitor going out of stock does not repeat. Growth that came from advertising spend repeats only if you spend again.
In 2026, Thanksgiving falls on 26 November, Black Friday on 27 November, and Cyber Monday on 30 November.
To have inbound stock checked in and sellable by 20 November, you need the full chain: ocean transit, port clearance, drayage, and the receiving lead time at the destination, which for marketplace fulfillment networks can add a week or more of its own during peak. Call it 45 days of transit plus 10 days on the ground.
Fifty five days before 20 November is 26 September 2026. That is the last realistic purchase order date for an ocean shipment intended to sell during Cyber Week this year. Anything later is an air freight decision, and air freight is a margin decision, not a logistics one.
Marketplace receiving cutoffs are published each year and they move. Check the current dates for each channel rather than assuming last year's held.
A restock report can compute most of this. The restock report inside ConnectBooks is driven by sales history and velocity, and it factors in supplier lead times and inbound stock already on the water, which are the two inputs sellers most often forget.
What it does not do is apply a seasonality index. That step is yours. Practically, that means you run the restock report, then multiply the recommended quantities for October through December orders by your own monthly index, then sanity check the result against the calendar above.
Being clear about that boundary is more useful than pretending otherwise. A seller who assumes the tool has already handled seasonality will underbuy, and will find out in the third week of November.
The inventory layer tracks stock by warehouse and holds transfers as in transit until received, which matters here: inbound units already counted as inbound should not be counted again when you compute the gap.
Step six is the one that gets skipped. A reorder point is not a setting. It is a calculation with a shelf life of about three weeks during a season that moves this fast.
The cost of being short is not the revenue you did not book. It is the contribution margin you did not earn, plus the ranking decay that lasts weeks past the outage, which is worked through in the true cost of stockouts. The cost of being long is carrying value and eventual markdown. Those two numbers are rarely symmetric, and knowing which one is larger for a given SKU is what profit reporting by SKU and channel is for. The rest of the seasonal preparation sits in the Q4 inventory readiness checklist, and the channel side runs through connections like Amazon accounting.
Running an e-commerce business comes with plenty of challenges, but ConnectBooks is here to make your life easier. With real-time insights, seamless integrations, and detailed tracking of your profitability and inventory, you can stay ahead of the game. Whether you’re selling on Amazon, Shopify, Walmart, TikTok or eBay, ConnectBooks helps you manage your finances with 100% accuracy and confidence, so you can focus on growing your business.
Ready to level up? Start making smarter, data-driven decisions every step of the way. Try ConnectBooks Free Today or Schedule a Demo
