Forecast Accuracy
Your Forecast Isn't Wrong. Your Process Is.
Merchandising Solves This | Week 4
Ask anyone who builds seasonal forecasts for a living and they'll tell you the same thing, usually with a tired laugh. The forecast is never right. Not exactly. The real question is how far off it ends up, and whether anyone learns anything from the misses before it happens again.
I hear a version of this constantly. "Our forecasts never match real-world sell-through." Said almost word for word, by brand after brand, retailer after retailer, like it's just the cost of doing business.
It isn't.
And when it feels that way, the problem usually isn't the forecast itself. It's the process that built it.
Where Forecasts Go Wrong
A forecast is only as good as what feeds it. Most of the time, that's some combination of last year's numbers, a growth target set above the planning team, and a fair amount of optimism from whoever is closest to the product.
What's usually missing is actual historical sell-through across channels, by season. Not just what shipped. What sold, at what price, and how fast, across ecomm, owned retail, and wholesale alike.
Wholesale shows shipped units long before it shows sold units, since sell-in and sell-through run on different timelines. Ecommerce and owned retail have their own version of this gap, full-price sales versus sales that only happened at a markdown. Across every channel, shipped, sold, and sold at full price are three different numbers, and the space between them is where most forecasting problems quietly start.
The Three Places the Process Breaks Down
The forecast starts from a target, not from demand. Leadership sets a growth number, and the math gets reverse engineered to land on it, rather than the goal being tested against what historical sell-through actually supports. That's not forecasting. That's math in service of a hope.
Sales optimism and production minimums quietly inflate the number. A sales lead wants enough inventory to never miss a sale. A factory has minimum order quantities that don't bend easily to a precise forecast. Both pressures push the number up independently, before anyone checks whether the combined total still makes sense against actual demand.
Last year's forecast becomes this year's baseline, growth percentage included. This treats last year's forecast as a fixed, accurate starting point, when that number may have already included its own overbuy or misses. If last year's forecast was off, building this year's directly on top of it just repeats the same miss with a new growth percentage attached. And when that pattern repeats across multiple seasons without real sell-through data to course correct, each forecast compounds the one before it, making the gap between what's planned and what's selling wider every year.
What Merchandising Brings to Forecasting
The strongest forecasts don't rely on bottoms-up planning only or top-down ambition alone. They need both. This is where merchandising plays a critical role, bridging the gap between what demand supports and where the business wants to grow.
Forecasting is a genuine collaboration. Sales, finance, product, marketing and merchandising all bring something real to the table. But it's merchandising that brings the product context no other function owns: which models are true heroes worth deepening, which colorways have proven staying power versus which are seasonal and should be planned as one-and-done, which products are linked by a marketing story and need to move together, which channels a product is right for and which it isn't. And let's not forget about launch timing.
That context is what turns a number in a spreadsheet into a real investment decision. Without it, planning can optimize a forecast technically and still be planning deeply into the wrong products.
In practice, that means building the forecast bottoms-up from real historical sell-through by model, across every channel, alongside the top-down growth ambition the business wants to pursue. Merchandising helps connect those two things instead of letting one quietly override the other. And it means closing the loop every season, comparing what was forecasted to what sold, model by model, channel by channel, and asking why the gap exists where it does and what can demand support going forward. Sometimes it's a true demand miss. Sometimes it's pricing, or a markdown masking weak full-price sell-through, or a product that wasn't differentiated enough to compete for the sale in the first place.
The Forecast Is a Hypothesis, Not a Promise
A forecast was never meant to be a guarantee. It's a hypothesis, built on the best available data, tested every season.
Treated that way, a miss isn't a failure. It's information. The brands that improve their forecasting over time aren't the ones who happen to guess better. They're the ones who go back, look at where the hypothesis broke down, and adjust the inputs for next time. And critically, they ask not just whether the numbers were off, but whether the product context feeding those numbers was as clear as it needed to be.
That's a merchandising question as much as a planning one. How well did the team understand which products deserved the deepest investment before the forecast was built? If the answer is "not well enough," that's where to start next season.
Where to Start
Pull last season's forecast next to what sold, model by model, for each channel of your business. Not in aggregate, since the aggregate can look fine while individual models or channels were wildly off in opposite directions that cancelled each other out.
Identify your biggest misses in both directions. Overperformance often reveals real demand that wasn't built into the forecast. It's worth understanding why it overperformed before deciding how to plan it next season, whether that means buying deeper, adjusting the price, or expanding the colorway run. Underperformance often points to a pricing, timing, differentiation, fit, or assortment depth issue.
Before building next season's forecast, align on the product hierarchy first. Which models are heroes? Which colorways are core versus seasonal? Which products are tied together by a marketing story that needs to be planned and supported as a group? Those answers should come from merchandising, and they should come before the numbers get built, not after.
Once that model-level foundation is solid, layer in colorway and size planning. Trying to plan that level of detail before the model-level demand picture is clear just compounds the same miss further down.
A forecast built this way won't be perfect. No forecast ever is. But it will be honest about what it's based on and built around the products that truly deserve the investment. That's what makes it improve season over season instead of repeating the same miss year after year.
The Outdoor Merchant is a product merchandising consultancy specializing in outdoor, cycling, and snowsport industries. Each week in this series, we explore a real business problem that smart merchandising was built to solve.
Follow along for Week 5, and reach out to The Outdoor Merchant if any of this is hitting close to home.