Demand forecasting is the discipline of turning what you already know into a buying plan with commercial logic behind it. It is one of the most useful skills in fashion buying and one of the least formally trained, which is why most buyers arrive at their unit numbers by a mixture of last season, their merchandiser’s suggestion and what felt right in the market.
Jump to:
What Demand Forecasting Actually Is
When you decide how much of a product to buy, how many units, across how many options, in what size ratio, where does that number come from?
For most buyers in the first few years the honest answer involves some version of what we did last season, or what my merchandiser suggested, or what felt right when I was walking the market. All of those are reasonable starting points. None of them is forecasting.
Demand forecasting is the practice of taking your sales data, your market research and your historical performance, and translating them into a plan that drives buying decisions with commercial reasoning behind them rather than instinct alone. It is not complicated. It is a discipline about how you use information you already hold.
Podcast Thrive in Fashion Buying & Merchandising
Episode 044, Forecasting Like a Pro: Bridge the Gap Between Gut Feel and Data.
Also on this topic:
Episode 016, The Fashion Buying Crystal Ball: How I Predict What Will Sell 6 Months From Now.
Forecasting is the skill that makes open to buy work as a planning tool and makes margin targets achievable during the buy rather than after it. The Thrive in Fashion Buying and Merchandising course teaches it as a connected process alongside OTB and margin, which is how it operates in a buying office.
Forecasting Is Not Trend Prediction
The two get used interchangeably, and the confusion costs buyers real money.
Trend prediction identifies what will be aesthetically relevant, what direction product should move in, what your customer will want to wear. It is a creative and market intelligence exercise and it matters. It is not forecasting.
Forecasting is about quantity and timing. How much of something will sell, when it will sell, and at what rate. You can be entirely right about the trend and still lose margin, because the direction was correct and the depth was guesswork.
Your Forecast Is Not the Sales Plan
The second thing worth separating is your forecast from the plan you were given.
The sales plan your merchandiser presents sets the financial target. Your forecast is your own view of how the products you are buying will perform against that target. Which lines drive volume, which underperform, where you need more depth, where you can afford to go narrower.
In practice, forecasting means asking questions before you commit. Based on last season’s performance on this product type, what sell-through should I plan for? If I buy 500 units and expect 65 per cent to sell at full price, what markdown am I budgeting for the remaining 35 per cent, and what does that do to my margin? If the answer is uncomfortable, do I need to reduce the depth of the buy?
This is also an in-season activity, not a pre-season one. Budgets and targets are set well before trading starts. Once you are in season, revisit your assumptions at every open to buy meeting and apply what the current trading pattern is telling you.
The Three Data Sources You Already Have
What stops most buyers forecasting is the belief that they need better data. In almost every case that is not true.
- Your own sales history. Last season’s performance by product type, category and price tier. Sell-through rates, average selling prices, markdown depth. This tells you what your customer actually bought when given the choice, not what you hoped she would buy. For every new decision, find the closest comparable product and look at how it performed.
- Your market context. What comparable retailers are doing with similar product, which price points are selling and which are stalling. This does not need to be formal research. History tells you what happened. Market context tells you whether the conditions that produced that result are likely to repeat.
- Your in-season trading data. Live information on performance against plan: which styles are running ahead of forecast, which are behind, where size ratios are off, which price points your customer is moving towards. This is where you chase into strong performance if the budget allows, or take early action on slow movers before markdown pressure builds.
Buyers who only look at data in a post-season review are working with half the picture. Historical analysis matters, and so does reforecasting while the season is still live and there is something you can do about it.
Where Instinct and Data Break Down Without Each Other
Instinct is not the problem. A good buyer’s instinct comes from years of market exposure and pattern recognition, and it is valuable. The problem is using instinct where data should be leading, or using data only to confirm what instinct has already decided.
Take range building. A buyer is genuinely excited about a trend direction and builds a significant proportion of the range around it. Her read on the direction is correct. But without a forecasting discipline, the buy depth on individual styles is set by enthusiasm rather than demand planning. The range performs. Three of the seven styles never sell through, and the markdown cost eats the margin the trend was supposed to deliver.
Or take a repeat order on a strong seller. The instinct that there is more opportunity in the style is often right. Without asking how much of that early sell-through is genuine ongoing demand rather than an opening burst, what the week by week rate of sale looks like, and whether the delivery window catches the remaining selling period, the reorder lands late or in the wrong quantity. The instinct was sound. There was no structure underneath it.
Instinct tells you direction. Forecasting tells you the quantity, the timing and the risk. You need both, and a buyer who has only one of them will be found out eventually.
Reading the Market as Part of the Forecast
Walking the market is where the two halves meet, and buyers do it differently to shoppers.
When I am comp shopping I look first at what meets me at the entrance, because that is the statement the retailer most wants to make. Then how much is new against how much is old, where the volume lines sit, and what they have clearly invested in. What I am looking for is repetition across stores, because if your customer is walking in and out of the same shops she is absorbing that repetition too, often without registering it.
Working in mass market, you also need to look a level above where you sit. By the time something reaches the high street it has already been diluted across every competitor. Looking at more directional designers or the catwalk gives you the raw version, and your job is the interpretation that is right for your customer.
A practical detail worth knowing: shapes do not run neatly from one season to the next. A silhouette can carry from autumn winter through spring summer and reappear the following autumn. Keeping it fresh across that span is a matter of new colours, alternative prints, changed trims, a long sleeve version of what was short sleeved. That is a forecasting decision as much as a design one, because it determines how much you can commit to a shape you have already sold once.
Elisabeth's Final Thoughts
Forecasting is the practical skill that makes everything else in buying work better. It is how open to buy becomes a planning tool rather than a limit, how margin targets get achieved during the buy rather than defended afterwards, and how your instincts get to operate inside a structure that protects them. With artificial intelligence now touching so much of the business there is a real risk of overanalysing, of having so much information that the decision gets harder rather than easier. Keep it simple, and go back to the three sources you already have.
Related reading
- What Is Open to Buy (OTB) in Fashion Buying?
- Fashion Range Planning: How to Maximise Sales and Margin
- Sell-Through Rate: Formula, Benchmarks and What It Tells You
- Your Guide to Seasonal Planning for Fashion Buyers
- How Fashion Buyers Actually Spot a Winning Product
- Fashion Forecasting: How Buyers Evaluate a Trend
Demand Forecasting FAQs
Demand forecasting is the discipline of translating sales history, market context and in-season trading data into a view of how much of a product will sell, when it will sell and at what rate. It sits behind every decision about buy depth, option count and size ratio, and it is what turns a sales target into a buying plan.
Trend forecasting identifies what will be aesthetically relevant and what direction product should take. Demand forecasting is about quantity and timing: how much will sell, when, and at what rate. They are related and trend research feeds into forecasting, but they answer different questions and require different evidence.
No. The sales plan sets the financial target you are working towards. Your forecast is your own view of how the products you are buying will perform against that target, which lines will drive volume, which will underperform, where you need more depth and where you can go narrower. One is the target, the other is your commercial position on it.
You almost certainly have enough already. Your own sales history by product type, category and price tier. Your market context, meaning what comparable retailers are doing and which price points are selling. Your in-season trading data showing performance against plan. The problem is rarely a lack of data. It is knowing what question to ask of it.
Ask how much of the early sell-through is genuine ongoing demand rather than an opening burst, look at the week by week rate of sale rather than the cumulative figure, and check whether the delivery window on a repeat is tight enough to land inside the remaining selling period. A repeat that arrives late is a repeat that becomes markdown.
