Seasonal demand planning in fashion: a phase-by-phase guide

By Elie Dufeu, CTO & Co-Founder, Metreecs. Published September 26, 2026.

Seasonal demand planning in fashion is the process of forecasting how much of each product, size, and color will sell in each store and channel over a season, then adjusting that plan as real sales arrive. It works best as four separate decisions (the pre-season buy, the first read, the in-season chase, and the exit) rather than one big bet placed months in advance.

The timing is what makes fashion hard. McKinsey notes that moving a product from design to market can take nearly a year, so most of the buy is committed long before a single customer has seen the collection. Planners and buyers who treat the season as a sequence of smaller, data-backed calls protect more full-price sales than those who try to get the opening order perfect.

Key Takeaways

• Fashion product development can take nearly a year from design to market (McKinsey), so the pre-season buy is always placed with the least information the team will ever have.
• 46% of fashion executives expect industry conditions to worsen in 2026, with tariffs cited as the number-one hurdle (McKinsey and The Business of Fashion), which makes committing 100% of the buy up front riskier than it used to be.
• Splitting the season into four decision points (pre-season, first read at weeks two to three, in-season chase, exit) gives the plan three chances to correct course instead of one.
• The first two to three weeks of sell-through are the strongest signal a planner gets, and forecasts at product x size x store level turn that signal into specific reorder and reallocation calls.
• How much of the open-to-buy budget to hold back for in-season reorders depends mainly on supplier lead time: a four-week nearshore supplier allows a far larger reserve than a 16-week offshore one.

What is seasonal demand planning in fashion?

Seasonal demand planning in fashion is the practice of forecasting demand for a defined selling window (spring/summer, fall/winter, holiday, or a capsule drop) at product, variant, and location level, then turning that forecast into buy quantities, store allocations, and in-season adjustments. The goal is to sell as much as possible at full price and exit the season with minimal leftover stock.

It differs from year-round demand planning in one structural way: most seasonal products have a fixed end date. A basic white shirt can be replenished for years. A printed summer dress has perhaps 12 to 16 weeks to sell before it becomes markdown stock. That fixed window changes every calculation, from safety stock to reorder point, because there is no “next cycle” in which to catch up.

Seasonal plans also carry more new products than a continuity range. A large share of a collection has no sales history at all, so the forecast has to borrow signal from comparable past styles, product attributes (category, price band, fabric, silhouette), and the early weeks of the season itself. If you want to see how a product-level forecast handles new styles like these, AI demand planning for seasonal collections builds the demand signal for each product and store before the first unit ships.

The four decision points of a fashion season

A season is easier to plan once it’s broken into the moments where a decision can still change the outcome. Each one runs on different data and has a different cost of being wrong.

Pre-season: planning the buy with the least information

The pre-season plan sets the total quantity per product and splits it by size, color, and store. Most teams start from the closest comparable season and adjust for known changes: new stores, a shifted price band, a category that grew last year.

Where this phase usually goes wrong is the distribution rather than the total: a buy that looks right at category level can still send the wrong size curve to half the stores, because size demand varies more by location than most planning spreadsheets account for.

The first read: weeks two to three

The first two to three weeks of sales are the most valuable data point in the season. Early sell-through by product and store tells the planner which styles the customer actually wants, often with enough confidence to act. Waiting for week six to “be sure” usually means the chase reorder arrives after the demand peak.

In-season chase and reallocation

Once winners and slow movers are visible, the plan shifts from forecasting to reacting. Strong products get a chase order if supplier lead time still allows one, while slow sellers move from low-demand stores to locations where the same product is selling through. Where contracts permit, late deliveries on weak styles are pushed back or cancelled.

Exit: planning the end of the season

The exit is where the earlier decisions show up in the margin. A season planned well needs fewer and shallower markdowns, because stock was already moved to where it sells. The practical questions are when to start markdowns, by how much, and in which locations.

How to build a seasonal demand plan, step by step

The steps below apply whether the range is womenswear, footwear, outerwear, or accessories. The detail level is what separates a plan that holds from one that gets rebuilt in week four.

1. Group new products with their closest analogs. For every new style, identify two or three past products with similar attributes (category, price band, fabric weight, fit) and use their sales curves as a starting point. This gives each new product a demand shape instead of a flat guess.
2. Forecast at product x size x store level, not category level. A category forecast can size the total dress buy, but it won’t show that a coastal store sells twice as many size S as an inland one. The plan needs to carry that detail from the start.
3. Set the size curve per store cluster. Group stores with similar size demand and apply a separate curve to each group. In our experience deploying store-level forecasting, a single national size curve is behind many of the broken size runs we see mid-season.
4. Decide how much of the budget to commit up front. Split the open-to-buy into an initial commitment and a reserve for in-season reorders. The split depends mostly on supplier lead time: a nearshore supplier with four-week lead times allows a larger reserve than an Asian supplier at 16 weeks.
5. Set safety stock for the season window, not the year. Seasonal safety stock should shrink as the season progresses, since the cost of excess stock rises as the end date approaches.
6. Define your first-read triggers before launch. Agree in advance which sell-through thresholds at week two or three will trigger a reorder, a transfer, or a hold. Decisions made in advance happen faster than decisions made in a meeting.
7. Replan weekly during the selling window. Update the forecast with actual sales every week, and let the plan flag the products and stores that have moved furthest from it.

Claire, head of planning at a mid-market womenswear retailer with 35 stores, rebuilt her fall plan this way after a season where the national size curve left 11 stores out of size M on the lead coat by week three, while larger sizes piled up in the same locations. Moving to three store clusters with separate size curves, and setting a week-two reorder trigger, cut the number of stores with broken size runs on her top 20 products from 11 to three in the following fall season.

Where seasonal demand planning breaks

Even well-built seasonal plans run into the same structural limits, and they come from the way seasonal data is shaped rather than from a lack of effort.

New products have no history. Analog and attribute-based forecasting helps, but the first forecast for a new style will always carry more uncertainty than a continuity product. The fix is to plan for correction rather than precision: smaller initial depth on the riskiest styles, faster first reads.

The signal arrives late. By the time sell-through is reliable, lead times may already rule out a reorder. This is why the budget split in step four matters so much. A plan with no reserve and long lead times can only react through transfers and markdowns.

Category-level plans hide store-level problems. A category can be on plan in total while individual stores sit at opposite extremes. The average looks healthy, and the problem only becomes visible once markdowns start. Inventory optimization at product and store level exists precisely to catch that imbalance while there’s still time to act on it.

External shocks move faster than seasonal calendars. Weather, tariffs, and shifts in consumer spending don’t wait for the next planning cycle. McKinsey’s State of Fashion 2026 report found that 46% of executives expect conditions to worsen this year, which argues for plans that can absorb a surprise mid-season rather than plans built on a single fixed scenario.

One mistake we repeatedly see is treating the pre-season buy as the plan and everything after it as damage control. The pre-season number is a starting hypothesis. Teams that plan the correction points as carefully as the initial buy tend to exit the season with cleaner stock positions.

Seasonal forecasting for fashion retail: pre-season vs in-season planning

Pre-season and in-season planning answer different questions, and the teams running them often sit in different parts of the business.

Good seasonal demand planning keeps both on one shared forecast. When pre-season planning lives in a buying spreadsheet and in-season planning lives in a separate allocation tool, the team loses the link between what was expected and what’s happening, and every variance has to be explained by hand.

Across Metreecs’ work with fashion and footwear retailers, the gap between these two teams is often where margin leaks. Buyers own the original quantities, allocators own the in-season moves, and nobody owns the question of whether the season is still on track at product and store level. For a deeper look at handling demand spikes inside the season, see the guide on managing seasonal peaks without overstocking.

Tomas, planning manager at a footwear chain with 50 doors, ran exactly this split for three seasons. When his team moved pre-season and in-season planning onto the same product x store forecast, the spring sandal range showed a clear week-two signal: two colorways were selling at roughly twice the planned rate in southern stores. Because the reorder trigger had been agreed before launch, the chase order went out that week, and the two colorways ended the season with 94% full-price sell-through instead of the 70% the team had budgeted.

Reducing end-of-season markdowns through better seasonal planning

Markdowns are the most visible cost of a seasonal plan that went off track, but they’re usually a symptom of decisions made weeks earlier. Most end-of-season excess can be traced to one of three moments: an initial buy that was too deep on a weak style, a size curve that didn’t match the store, or a slow-selling product that stayed in the wrong location too long.

A few practices reduce markdown exposure without changing the buying budget:

• Move stock before you discount it. A product selling slowly in one store is often selling well in another. Transfers at week four or five cost less than a 30% markdown at week ten. The guide to smarter stock allocation for full-price sales covers how to prioritize those moves.
• Plan markdowns by location, not chain-wide. A product with eight weeks of cover in one store and two weeks in another doesn’t need the same price in both.
• Track weeks of cover alongside sell-through. Weeks of cover at the current sales pace shows whether the remaining stock will clear before the season ends, which a backward-looking sell-through rate can’t answer on its own.
• Start small and early. A shallow markdown at the right moment often clears more stock than a deep one applied late.

Amara, merchandise planner for an accessories brand selling through 20 stores and its own e-commerce site, applied location-level markdowns to her summer bag range for the first time last year. Instead of a chain-wide 40% cut in week ten, she ran 20% markdowns in the six stores with the highest weeks of cover from week eight, and transferred remaining stock from those stores to the online channel. The range closed the season with 18% less leftover stock than the previous summer, on a similar buy.

Frequently asked questions

How do fashion retailers plan for seasonal demand? Most start from comparable past seasons and adjust for known changes like new stores, price shifts, or category trends. The stronger approach splits the season into a pre-season buy, an early read of sell-through in weeks two to three, in-season reorders and transfers, and a planned exit. Each phase uses the most recent data available at that point.

How do you forecast demand for a new fashion product with no sales history? Use analog forecasting: find two or three past products with similar attributes (category, price band, fabric, fit) and use their sales curves as a starting shape. Then correct that forecast quickly once the first weeks of real sales arrive, since early sell-through is more reliable than any pre-season assumption.

How much of the seasonal budget should be committed before the season starts? It depends mainly on supplier lead times. With long lead times, most of the budget has to be committed up front, and in-season flexibility comes from transfers and markdown timing. With shorter or nearshore lead times, retailers can hold back a larger reserve for chase orders on the products that prove themselves early.

What is the difference between pre-season and in-season planning? Pre-season planning decides how much to buy and where to send it, using historical data and assumptions. In-season planning adjusts that plan once real sales come in, through reorders, stock transfers between stores, and changes to deliveries. Both work best when they run on the same forecast.

How can I reduce end-of-season markdowns in fashion? Most markdowns trace back to earlier decisions, so the biggest gains come from better size curves per store, faster reactions to early sell-through, and moving slow stock to stores where it sells before discounting it. When markdowns are needed, planning them by location and starting earlier with shallower cuts usually clears more stock at a better margin.

Conclusion

Seasonal demand planning in fashion works when the season is treated as a series of decisions rather than a single buy: plan the pre-season depth carefully, read the first weeks fast, chase and reallocate while there’s still time, and exit with markdowns planned by location. The common thread is a forecast detailed enough, at product, size, and store level, to make each of those calls with confidence.

Book a session with our team to see how a product x store forecast supports each decision point of your next season.

Sources

• McKinsey & Company, “Fashion on demand” (from The State of Fashion 2019): design-to-market timelines of nearly a year. https://www.mckinsey.com/industries/retail/our-insights/fashion-on-demand
• McKinsey & Company and The Business of Fashion, “The State of Fashion 2026: When the rules change”: 46% of executives expect conditions to worsen in 2026, tariffs cited as the top hurdle. https://www.mckinsey.com/industries/retail/our-insights/state-of-fashion