The question "how do I calculate a seasonal supplier order" lands on every buyer's desk at least twice a year — right before the demand peak and right after it. Overshoot, and your cash is frozen in dead stock that later has to be cleared at a discount. Undershoot, and you get empty shelves in your most profitable month while customers walk over to a competitor. Yet the calculation itself is neither magic nor gut feel: it comes down to a few data sources and one formula with clear guardrails. In this article we'll walk through that logic using the example of an import wholesale business with dozens of brands and thousands of SKUs.
Why ordering "by eye" produces two extremes
"By eye" is when the buyer opens the supplier's price list and fills in quantities from memory: we sold a lot of this, less of that, this one's new, let's try a few. The problem isn't the person's competence. It's scale: when you're running dozens of brands and thousands of SKUs, no memory can physically hold the sales history for every line.
The result is almost always one of two things:
- Dead stock. Items ordered "to be safe" sit for months and years. Cash that could be working in your turnover is frozen in boxes on the shelf.
- Empty shelves in season. A fast mover wasn't reordered, the peak arrives, and there's nothing to sell. For an importer this hurts even more: you can't reorder quickly, since the shipment takes months to arrive.
- Skewed assortment. Familiar items get ordered generously, while less visible ones get forgotten — even though those were the ones quietly delivering steady margin.
Both extremes come from the same cause: the decision is made without data the business actually has. Sales history sits in your accounting system, stock is on the shelf, prices are in supplier price lists. All that's left is to pull it together and do the math.
What data you need before you calculate
Before you can calculate anything, you need to gather four sets of data. Miss any one of them and the calculation turns into guesswork.
- At least 13 months of sales history. Why 13 and not 12: it puts the same month from last year into your sample. If you're placing an order in July, you can see all of last July too — and that's the foundation of a seasonal forecast.
- Current stock on hand. Free stock that isn't already reserved for customers.
- Goods in transit. For an importer this is critical: a shipment can be on the water for six to eight weeks, and if you don't account for what's already ordered, you'll order the same thing twice.
- Purchase prices. So you know how much cash the order will tie up, and so you can check new price lists against old ones. We covered how to build a price history by supplier in the article on tracking purchase prices.
The good news: all of this already exists in any working business. The bad news: the data is usually scattered across the accounting system, Excel files, and email, and before every order someone spends a day stitching spreadsheets together.
ABC analysis of purchasing: not all items are equal
ABC analysis of purchasing means splitting your assortment into three groups by their contribution to turnover. The classic logic: group A is a small share of items that drives the bulk of revenue; group B is the middle; group C is the long tail of items, each selling only occasionally.
Why a buyer should care:
- Group A — the zero-tolerance zone for stockouts. These items must always be in stock, and a safety stock on them is justified: the cash you tie up is less than what you lose from an empty shelf.
- Group B — standard formula-based calculation, without hand-holding every line.
- Group C — candidates for made-to-order buying or rare batches. Holding safety stock across the whole tail is a reliable way to grow dead stock.
The practical point: ABC analysis doesn't answer "how much to order," it answers "where to look closely." A buyer physically can't think through thousands of lines — but they can think through the few hundred lines in group A and trust the formula with the rest. It's important to recalculate the groups regularly: items migrate between them, and last year's classification eventually stops reflecting reality.
Purchase planning around seasonality
Purchase planning around seasonality answers one question: how does this month differ from the average. For most products, demand is uneven across the year, and a single average figure hides that. In heavy-truck parts, for example, cooling-system components move in summer, while everything tied to engine starting moves in winter.
The basic way to gauge seasonality is to look at the same month last year and the months around it. If last August ran noticeably above the annual average for an item, it's reasonable to expect a similar picture this August — provided the item was actually in stock back then. An important caveat here: if the item ran out mid-peak last season, the sales history understates real demand, and you can't copy it blindly.
For an importer, seasonality stacks on top of lead time. If a shipment takes two months, an order for the August peak has to be placed in May or June. Suppliers know this and often send pre-season offers with deadlines and special terms — missing one of those emails in the daily flood is expensive. On how important emails drown in the inbox and what to do about it, there's a separate breakdown.
How to calculate a seasonal supplier order: the working formula
Let's pull it all into one formula. In the import wholesale business whose logic this breakdown is built on, the per-item calculation looks like this:
Order = MAX(seasonal demand; 13-month average) × 1.4 − Available
Let's break it down piece by piece:
- Seasonal demand — the sales forecast for the period this order will cover, adjusted by the seasonal factor from the same month last year.
- 13-month average — insurance against anomalies. If last season was distorted by a stockout or a one-off large deal, you take the larger of the two numbers: the seasonal forecast or the stable average. That way the formula won't let the order "collapse" because of one bad data point in the history.
- The 1.4 factor — a safety margin for demand growth and delivery delays. It's not a universal constant but a setting for your specific business: it depends on your lead times and the cost of a stockout. What matters is that the factor is the same for all items and set deliberately, not eyeballed line by line.
- Available — free stock on hand plus goods in transit. You subtract everything you already have or that's already on the way.
If the result is negative or close to zero, you leave the item out of the order.
Guardrails: the two-month rule and common sense
The formula is the baseline, but it has blind spots. That's why, on top of the calculation, the same business runs hard guardrail rules.
- The two-month rule. Don't order anything you already hold more than two months of sales of. Not even if the formula computed something, not even if the supplier is running a promotion. Two months of stock plus goods in transit is already a sufficient cushion; anything beyond it is frozen cash.
- Item life cycle. If a part belongs to a model range that's being phased out, the sales history lies: it shows past demand that soon won't exist. Such items go in only against confirmed demand.
- Price check. Before placing an order, the new price list is checked against the old one. A deviation within reason is normal market movement. A difference of several times over is almost always an error in the price list or in the units of measure — and it's cheaper to catch before the order than after the invoice.
- Rhythm. The order for each supplier is placed on a regular cadence — say, once a month — rather than "whenever someone remembers." Regularity alone reduces both stockouts and impulse buying.
How much to order: the system calculates, the buyer decides
The right division of labor looks like this: the system prepares the per-item calculation, the human makes the decision. Not the other way around.
Answering "how much to order" across thousands of lines is mechanical data work: pull the history, apply the formula, cut items by the two-month rule, tag the ABC groups. You can do it by hand in Excel, but that's a day or two of work before every order, and errors in manual spreadsheets are inevitable.
And here's what a machine doesn't know and can't know:
- the supplier has announced a price increase from next quarter — worth taking more now;
- a large customer has planned a project — you need stock beyond the statistics to cover it;
- an alternative brand has appeared — some items are better replaced than reordered.
So the output of the system isn't a sent order but a payment draft for the order, with a calculation and an explanation for every line: why exactly this quantity. The buyer walks through the draft, adjusts the debatable lines, and confirms. Responsibility stays with the human — but for the first time they have complete data to decide on.
How this works in HORUVIA
HORUVIA is an AI operator for businesses that live in email and documents. Order recommendation is one of its functions, and it's built on exactly the logic described above: seasonality, ABC analysis, and current stock on hand.
Here's how it fits into the working loop:
- Supplier memory. The system parses supplier email, sorts price lists and invoices by brand, and accumulates a price history — so that by order time you have both current prices and how they've moved.
- Order recommendation. By the order date the system prepares a draft using the seasonality-and-ABC formula, subtracts stock on hand and goods in transit, and cuts items by the two-month rule.
- Telegram alerts. An offer arrives with a pre-season order deadline — the buyer hears about it right away, not after the deadline has already passed.
- The decision stays with the human. The system doesn't send orders on its own. It calculates and explains the math; the buyer confirms.
Order recommendation is part of a larger loop: from parsing email to payment drafts and deadline tracking. For more on how these parts fit together, see the article on wholesale automation.
If you're calculating orders by hand today and want to see how the same logic works on your data, request a demo on the HORUVIA home page or estimate the buyer's time savings in the calculator. Answers to common questions about connecting email and data security are collected in the FAQ.