Seasonal inventory management: riding demand you can't fully predict

Seasonal inventory management is three problems stacked on top of each other: spotting which SKUs actually have a season, buying enough before the season starts, and getting out of last season's stock without torching margin. Small stores usually fail at the first one silently — they buy "extra for the season" on vibes, then discover the SKU was never seasonal at all, or was seasonal in the opposite direction. The fix is a two-year comparison on your own data, a buy sized to a season plan rather than a hunch, and a carryover decision made on the calendar, not in a panic.

Step 1: Detect seasonality in your own data

The honest way to find seasonality is the two-year comparison: pull monthly unit sales per SKU for the last two years, side by side. For each month, compute the ratio of that month's sales to the SKU's own average month.

Seasonality index = (month's unit sales ÷ SKU's average monthly unit sales)

A ratio of 1.0 means average. A December index of 2.4 means December sells 2.4× your average month — that's a season. An index that never leaves the 0.8–1.2 band means the SKU has no meaningful season, and treating it as seasonal just adds noise to your buying.

Worked example with hiking socks. Suppose your SOCK-HIK-MRN-M (merino hiking sock, men's medium) sells an average of 90 units/month across the year:

That's a fall-hiking season peaking in November — not a holiday-gift season, which would peak in December. The distinction matters: your buy needs to land by September, and a December promo would be aimed at the tail, not the peak.

Three honest caveats about the method:

• One year of data is not seasonality — it's one data point per month. With two years, you can average the indexes; with three, you can see whether the shape is stable. Under two years, treat your index as a rough draft and keep safety margins wide.
• Out-of-stock months corrupt the index. A SKU that sold only 30 units in November because it went out of stock on the 4th doesn't have a weak November — it has missing data. When you compute the ratio, either exclude or correct months where the SKU wasn't available the whole time. If a SKU was dark for part of a month, treat that month's demand as understated, not low.
• Growth distorts year-over-year comparisons. If the store grew 30%, last December was "bigger" relative to last year's baseline than it looks against this year's average. Compute each year's indexes separately, then average the shapes.

Step 2: Buy ahead of the season without overbuying

Anticipatory buying — ordering months before demand arrives — is where seasonal inventory gets expensive in both directions. Order too little and you're dark during the peak weeks that pay for the whole season; order too much and you're paying to store the mistake until next year.

The workable method for a small store: build the season plan as a cumulative curve, not one big number. Using the sock example, suppose you plan next fall around an indexed demand shape of Sep 1.33 / Oct 1.67 / Nov 1.78 / Dec 1.39, and you expect an average of 95 units/month (say the brand is growing). Your cumulative demand from September 1 through the season:

Sep: 126 · Oct: 159 · Nov: 169 · Dec: 132 · Total ≈ 586 units

Your pre-season order has to cover demand from order placement through the next replenishment opportunity. If your supplier's lead time is 45 days, whatever you order in July must cover September and most of October. The discipline is to place the pre-season buy sized to the first two months of the curve plus a safety buffer, and keep later months for a reorder informed by actual early-season sell-through — which is covered alongside the metric in our sell-through rate guide.

What this looks like as a decision, using illustrative numbers: order 290 units for the pre-season PO (September + October demand ≈ 285, rounded to the case pack), and hold a reorder decision for October 1 based on how September actually sold. If September came in at index 1.6 instead of 1.33, the reorder catches it. If September flopped, you haven't paid for a November and December you now have to discount.

Two structural points that decide whether this works:

• The reorder option only exists if the supplier can deliver mid-season. Ask before the season: can you do a second PO at 40–60% of the first, delivered in 3–4 weeks? If yes, you can split the risk. If no (long lead times, made-to-order, or minimums that force one big commitment), the pre-season buy has to carry the whole season, and safety stock math — not vibes — sets the buffer. The mechanics are in our safety stock guide.
• Anticipatory stock is a cash decision, not just a units decision. Stock bought in July sells in October; that's three-plus months of capital doing nothing. The cost of carrying that stock — roughly 15–30% of its value per year when you add storage, capital, shrinkage, and obsolescence, with your own rate as an assumption to work out — is covered with its formula in our carrying cost guide. A seasonal buy that earns its margin but only after four months of holding is a very different deal than the same buy with a two-week lead time.

Step 3: The end-of-season decision — carry over or clear

Somewhere in the last weeks of every season, each seasonal SKU needs one of two verdicts: carry it over to next year or move it out. The frame for the decision is a straightforward comparison:

Carry over if: expected full-price sales next season × margin > (clearance recovery now + carrying cost for the holding period)

Concretely: a seasonal SKU at a $6 unit cost and $16 retail with 300 units left. If you expect to sell 250 of them at full price next season, carrying 300 units for ~9 months costs (using an assumed 20%/yr rate) roughly $6 × 300 × 15% ≈ $270 in storage-plus-capital terms, and you'd still have 50 stale units to move. Carrying over probably wins. If you expect to sell only 80 next season, you're paying to hold dead inventory for a year — clearing now at a discount recovers cash you can redeploy into next season's buy.

The judgment calls that tilt the decision:

• Perishability of the product itself (dated goods, expiring materials) removes the carry-over option entirely.
• Fashion/content sensitivity — items tied to a trend, a year, or a design cycle lose value while sitting; evergreen basics don't.
• Storage reality — a garage-based operation holding 20 boxes of seasonal product for nine months pays in space and mental load even when the spreadsheet says carry.

One boundary to be clear about: this article covers the decision. If the verdict is "clear," the actual liquidation playbook — discounts, bundles, write-off sequencing — is a separate craft covered in our guide to getting rid of dead stock.

A calendar anchor helps more than any formula: set a fixed carryover review date each year — two weeks after each season's peak ends, for example — and run the comparison then. Decisions made on a date get made; decisions made "sometime after the holidays" get made in February, at worse prices, with less cash.

Special case: Q4

The fourth quarter compresses everything in this article — the biggest season, the longest pre-buy, the tightest reorder windows — and it has its own planning cadence with week-by-week deadlines that we won't duplicate here. If Q4 is approaching, our BFCM inventory planning guide owns that countdown.

Common questions

What is seasonal inventory? Seasonal inventory is stock whose demand concentrates in a predictable period of the year — holiday gift goods, back-to-school items, summer outdoor products. The test is an index above roughly 1.3 in the peak months versus the SKU's own average, consistent across years. Stock without that shape doesn't need seasonal buying logic; a flat reorder process handles it.

How do I know if my products are seasonal? Compare monthly unit sales over at least two years against the SKU's own monthly average. Months consistently above 1.3× average mark the season; months that never leave the 0.8–1.2 band mean no season. Correct for any months where the product was out of stock, because unavailable months understate demand and can fake a weak season.

How much seasonal inventory should I buy? Size the pre-season order to cover the first portion of the season's cumulative demand curve plus a safety buffer, and hold the rest for a mid-season reorder if your supplier can deliver one. If the supplier requires one big commitment, the safety stock calculation sets the buffer. Never size the whole season from a single average-month estimate.

What should I do with leftover seasonal inventory? Compare expected full-price sales next season against clearance recovery now plus the months of carrying cost in between. Carrying over suits evergreen, non-perishable items you're confident will sell; clearing suits dated, fragile, or weakly-demanded stock. Make the call on a fixed calendar date shortly after the season ends rather than letting leftovers drift into the year.

How is seasonal inventory different from safety stock? Safety stock buffers normal week-to-week variability and is sized by formula; seasonal (anticipatory) inventory builds months before demand and is sized by the season plan. They compound: a seasonal SKU usually needs both, because the peak months also carry the most variable demand. The buffer formula and its inputs are covered in our safety stock guide.

Next step

Pick your five SKUs with the most obvious season and build the two-year index table for them this week — it's an hour in a spreadsheet, and it will tell you whether your instinct about seasonality matches your data. Then check what your lead times imply for that season's order date, using the days-of-cover method in our days of cover guide.