For many fashion retailers much of the work towards ensuring a successful season will have already been done months before the collections go on the racks – stock for a summer season might well have been ordered almost immediately the previous summer season had ended, and in almost every case many weeks before a season begins. The essential tasks can broadly be broken down into two steps which this chapter will discuss more in detail.
Forecasting and Managing Seasonal Replenishment
Arriving at a forecast for a forthcoming season is often quite a time- and labor-intensive task where experience, and subjective opinion, play a much bigger role than objective, data based forecasts. In fashion, where novelty is so important and there are often no clear past examples on which to base future forecasts, this is understandable. The overriding importance of trends inevitably also means a high degree of manual planning. However, in our experience, statistical forecasting based on sales from previous seasons and similar products can provide a good baseline in planning for new seasons. In planning for new products it is possible to assign reference products automatically from previous seasons, which allows you to make an initial forecast for them, making it easier to add an expert view to those forecasts.
Often getting the prepacks and initial allocations right has a much bigger influence on total sales and margins than even the best-planned markdowns.
In fashion retail it’s not simply about sales figures for a given product – at some point you need to take that product-level figure and determine the optimal style to size-color split to reflect what consumers will actually buy. The best forecasts typically come from building a good product-level forecast bottom-up, applying market intelligence and budgetary targets to the top-level figures, and then splitting the forecasts into style, sizes, and colors based on, for example, store-level profiles, which can be derived from previous seasons’ data. Optimizing that style-split to sizes and colors can have a huge impact on the total sales as well as minimizing markdowns.
Successful Order Automation and Building Optimal Prepacks
Once an accurate forecast is created, then the rest is easy, right? If you’re lucky it can be relatively straightforward, but for most fashion retailers, the complexity continues. For lower-priced items, logistics and other processing costs mean that handling single units is just not economical and to cut those costs companies are forced to use prepacks: containing for instance one S, two M’s, three L’s, and an XL in a pack bundled together.
Building prepacks can be a bit of a nightmare, but it can have a huge impact on the bottom line: getting it wrong can mean that, despite having accurate forecasts, you still end up having too many L’s in one store and not enough in another. Often getting the prepacks and initial allocations (more on initial allocation in the next section) right has a much bigger influence on total sales and margins than even the best-planned markdowns
Four Steps to Optimize Prepack Creation:
- Build the store-product-size-color level forecasts for the season. This is the starting point – note that it’s vital to do it at store-level as prepacks are allocated directly to stores. So in building prepacks you need to match them to store-level forecasts.
- Understand all the constraints, once you have locked the forecast: how are you optimizing prepacks in terms of style to size-color split? How many different packs is it possible to build? Can you combine loose items with prepacks to give yourself more control?
- Manage the complex maths: determining the best number of different packs and optimal pack structures for each pack (so how many of each variant should go into a pack). As the number of different alternatives and combinations here is almost unlimited it is understandably very challenging to come up with the best option without the help of modern technologies.
- Enjoy the “recipe” for successful prepacks and the numbers of each pack needed.