Fit for Fashion

Fashion retailers need a more intelligent solution to size allocation, says David Hawkings, SVP of EMEA at antuit.ai (now part of Zebra), because they are dealing with a rising tide of markdowns and stock disposal issues in the age of sustainability.

The retailer that gets apparel and footwear size allocation per channel right is onto a winner in more ways than one. Complete stock sell through at full or near full price is the hard commercial goal, while the apparent soft benefits of customer satisfaction also convert to profit through repeat visits and higher transaction values.

The history of efficient size allocation is long and that is probably its biggest drawback. While right sizing has become more and more important to consumers, the methods used to get it right are often straight out of history as well.

Right now, many retailers are using a hybrid of new tech, even at the cutting edge of artificial intelligence, but basing all these technologies on old order management platforms, historical demand spreadsheets and all the instincts based around those.

The modern consumer is a lot less patient than they used to be and will happily quit shopping with a retailer that simply cannot get their sizes right. And this is a consumer that is now informed with a wealth of sizing advice from the media and via social media. In addition, the traditional size ratios simply do not match up in countries with growing ethnic, gender, occupation and income diversity.

This increasing diversity becomes even more challenging as retailers sell through more and more channels. Current allocation for style/colour/size do not allow for significant differences in size profiles between neighbouring stores and between countries. Between eCommerce and store the problem gets much worse because consumers shop sizes quite differently across these two channels.

So, the compulsion to act more intelligently on size is primarily commercial, but now it needs to consider sustainability and the commitments that the fashion industry has made to generate less waste, and also becoming carbon net zero. In addition, retailers will need to think harder about efficient allocation in stores as falling shopper traffic stresses current sales per square foot models.

It doesn’t take much to end up with a bad result; current models for margin loss show that the lost margin adds up quickly with a relatively small percentage of misallocated size units. Working with retailers, Antuit found that in a typical six-sized size range, misallocations of less than 20% of an order's units can drop margin dollars by as much as 50% if markdown pricing is uniform across all sizes. Additionally, lost sales quickly accumulate as customers cannot find their size. A missed allocation of 17% of ordered units yields up to 23% fewer sales as stock-outs depress demand.

In trying to solve these problems, there are many inefficiencies in what happens currently, mainly that poor demand forecasts at size level result in poor sales planning, which in turn leads to over-ordering of fringe sizes at the cost of lost sales for core sizes. Simplistic size profiling based on previous years cannot handle new and additional sizes, while new products and suppliers without a selling history cannot be modelled. This then results in retailers being unable to manage core business rules, such as the allocation of core sizes and coverage minimums. One consequence of this is that they over-compensate on final order quantities, particularly when they add perceived sales opportunities and pack constraints.

Efficient size allocation first needs a single management strategy, not a series of point solutions that can only solve elements. This can be achieved by developing demand forecasting that can spot an accurate selling signal among large volumes of selling data, and continuously balance service level (being in stock on needed sizes) with the cost of holding inventory.

Inside this strategy, retailers need to address certain realities - notably data-sparsity - resulting from imperfect data history, cleanliness and access, as well as a lack of history arising from the one-time, seasonal nature of fashion items, and also the difficulty of managing demand transference from an existing size set to a new one.

Solving data sparsity requires the creation of profiles at every level of a custom sizing hierarchy that includes product levels and attributes; analysing profiles at each hierarchy level for data sufficiency, and then assigning a weighting based on the sparsity of data at each level for the profile; and, possibly a composite size profile created by combining the multiple hierarchy profiles according to weight. This process provides detailed level profiles where selling volumes are sufficient, and where it isn't, profiles are inherited from higher levels.

To ensure items are in stock, taking a straightforward approach to coverage minimums pays dividends. By establishing a threshold value, by store, for what defines a core size and then defining the unit coverage minimum for the core sizes, retailers can apply coverage minimum responsive to store/product selling without the complexity of detailed rules sets.

Omnichannel must be natively built-in with the rest of the application, not as an additional add-on or +/- logic external to all decisions. Profiles should reflect store and for online sales independently as size demand differs significantly based on channel. Hence, solutions should not aggregate or intermix channel demand so that the distinctive size profiles are preserved.

Size profiling demand should also allow for ship-from-store orders, enabling any given location size profile to represent two demand sources: native sales and online sales fulfilled from store stock. This aspect becomes even more crucial as retailers fulfil more online orders from store.

Lastly, size profiling must integrate with any order management system (OMS). High-performance optimization allows quick API-driven size order quantities to be available as users write orders. Any system of value that provides intelligence within the current OMS infrastructure should enable a retailer to enjoy the advantage of advanced analytic determination of order quantities without replacing expensive ordering systems.

antuit.ai – now part of Zebra Technologies – is rethinking the way retail companies use AI, from supply chain to merchandising to marketing, to digitally transform their businesses to achieve substantial business results.