There is no shortage of fashion houses and retailers embracing AI to develop new ranges: Notably, H&M is using it to create the “digital dress”, Tommy Hilfiger is exploring “how AI can identify upcoming trends faster than industry insiders to enhance the design process”, Levi’s is using a virtual stylist chatbot to help customers choose the perfect pair of jeans and ASOS is using AI to recommend clothing sizes to shoppers based on what they’ve purchased and, crucially, kept in the past - dramatically reducing returns.
But despite compelling evidence, all but a few forward-thinking fashion brands are yet to reap the rewards.
The fashion merchandising chasm
Online merchandising, carried out manually, can never deliver the sales growth required to justify fashion retailers’ huge investment in driving quality traffic to their online stores.
Quite simply there are not enough available merchandisers, nor hours in the day, to effectively analyse the vast amounts of customer and product data required to optimise product exposure across catalogues running to tens of thousands of SKUs - much less, to do so in real time.
And there is no reason to believe that the old approaches will drive conversion rates above single figures - where they have stubbornly remained for years.
Everything from largely flat sales to inefficient use of resources, to the all-important sales and profit figures makes plain that a new approach is well overdue.
A solution hiding in plain sight
The only way to truly transform online fashion retail performance is to re-imagine merchandising for the online channel. It’s time to stop trying to shoehorn cumbersome, slow and ill-fitting offline techniques and turn to the power of AI and automation.
Marketing departments are spending more online than all other channels combined to drive qualified traffic to their websites. In the process they are embracing AI in the form of machine learning and predictive analytics, to automatically sniff out and signpost the most likely converting customers.
It’s baffling that once those marketers have delivered that quality traffic, there is a reluctance to use the same techniques where it really matters - in the store.
The sad truth is many merchandisers are simply too concerned about handing over the heavy lifting to a computer – they crave control.
Fashion merchandising, AI and conversion optimisation
The idea that fashion retailers cannot apply AI and automation to online merchandising is nothing short of nonsense. The fact is that the technology and tools required to do for online merchandising what Google is doing for PPC ads already exist - and the results are just as impressive.
For instance, an A/B test by Nordic beauty retailer Kicks pitched two versions of a homepage against each other - one manually produced by merchandisers using their experience and knowledge; the other completely delivered using AI and machine learning.
It came as no surprise to me, to see the automated page produce an eight times higher click-through rate than the manual version, eight times more add to basket events and, where the manual version generated one sale transaction, the automated version succeeded in generating 101 sales transactions.
Coincidence? Not likely.
It is clear that, if you can combine the great results from traffic generation and the results that AI merchandised websites deliver, there is an opportunity not only to take the horses to water, but to allow them to drink all they want. It’s time to let go.
Let go and grow, or hold on and sink?
But even though the commercial case for automated merchandising is compelling and inarguable, perhaps merchandiser’s reluctance to let go and hand over the heavy lifting to a computer is not such a surprise.
After all, many are too busy poring over reams of data to look around and see how the discipline is changing. Add to that the idea that handing over the reins to a machine means an end to the grand tradition of visual merchandising - it doesn’t - and misplaced doubt over a computer’s taste in apparel for instance, and the fact that automated merchandising remains the preserve of a forward thinking few, is entirely predictable.
Online fashion merchandisers are simply having difficulty letting go of what they know and feel safe with. But let go they must, because handing over to the machine means more time on strategic activity - freedom from all those lines of data - greater efficiency and more rewarding day to day jobs. Not to mention more sales, more profit and the chance to finally realise online fashion retail’s full potential.
So, the question is not whether AI can transform online merchandising results; it is which fashion retailers will be the next to make the leap, to let go and grow? They will gain not only transformational results, but lasting competitive advantage as they sell more while delighting and engaging customers.
So now it’s time to complete the circle: the machine can not only help you predict fashion trends and design apparel to suit, but ensure those digital designs are shown to the right shoppers in the right place at the right time.