Retailers need to reimagine merchandising for the online world - harnessing the power of artificial intelligence (AI) to transform the customer experience with real time relevance - if they are to realise business strategies and operational efficiencies to deliver much needed profitability.

The list of retailers announcing online growth strategies to escape the carnage on the high street grows almost daily. Many more see e-commerce innovation as the key to long term survival and growth.

Those retailers are responding to a number of factors - the high cost of premium high street real estate, falling in-store sales, and changing consumer behaviour to name but three - but a shift to online, as it is today, misses one very important fact.

Online retail conversion rates are not very good and are not improving much either. In fact, average rates have risen by less than 1% in five years, and now stand at less than 4.5%. To put that in context, offline stores average a sale for every four or five visitors - a conversion rate of around 25%.


Download a PDF version of this post for printing or reading offline

We respect your privacy. In our Integrity Policy you can read how we handle the information you provide.


What merchandisers know but aren’t saying

There is a simple, inescapable truth behind those stubborn conversion numbers. Online retail is a totally different animal to in-store:

  • It’s faster: Success depends on the ability to respond to trends in a heartbeat.
  • It’s about customer experience as much as product: Online it’s a split second click to go elsewhere, so satisfying, relevant user experiences are vital.
  • It’s not about the shop window: Every page is a potential landing page, so every page, every product must be optimised.
  • It’s labour-intensive: Merchandising catalogues of thousands of products and maintaining real time relevance is practically impossible if carried out manually.

And yet, the vast majority of retailers are still applying offline techniques to online stores - shop window merchandising, seasonal campaigns and flash sales. Ask most merchandisers and they will tell you this approach simply does not work online - and this, more than anything, is the reason for the sloth-like rise in conversion rates.

Rethinking merchandising for e-commerce

In fact, if it is to deliver transformative results, online merchandising needs a revolution. 

It is time to completely rethink merchandising for the online world. Long term success online depends on developing an approach that makes the most of talented merchandisers by focusing them on the big picture, and finding a new way to deal with the day to day ‘grunt work’ - the crucial, data-driven detail of predicting and adapting to consumer behaviour in real time.

The good news is this is already possible. The technology required to analyse millions of points of customer data, predict behaviour and adapt every aspect of the site to optimise sales in real time is already available. 

It’s just that CEOs don’t know about it yet - at least not in the context of online merchandising, even though it has already proven to transform results elsewhere.

AI for online retail

Artificial intelligence movies - from the Matrix to Ex Machina - have given AI a bad name. But AI isn’t here to subjugate humanity. Rather, we are harnessing the power of AI and machine learning to transform a whole series of industries, including online retail. But it has not yet been given the central role where its full power can be harnessed.

Marketing departments 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 taken the horses to water, their merchandising colleagues are limiting the amount of water they are allowed to drink. Many merchandisers are just too concerned about handing over the heavy lifting to a computer.


Why AI based merchandising?

Many e-commerce search and merchandising solutions claiming to employ AI are, in fact, rules based. They require the merchandising team to set up triggers based on criteria defined by poring over spreadsheets and other data sources.

The main issue with most merchandiser-created rules is that the trigger criteria is too broad, targeting too many customer journeys. Besides, rules can conflict, they are not truly scalable, they do not react in real time and creating them requires a lot of manual effort – back to square one.

Sophisticated AI software is self-learning, self-operating, consistent, accurate, scalable and efficient. It will run across all aspects of the site, learning and applying that knowledge to all facets. In fact, as data accumulates it just gets better and better.

This kind of data driven solution should enable retailers to hand over the heavy lifting of merchandising to computers, to put the right product in the right place at the right time, leaving them to set their overarching goals.

Conversion optimization - AI vs. humans

So, the technology and tools exist to do for online merchandising what Google is doing for PPC ads - but CEOs don’t know it yet. And if they did, and they knew just how potent the results are, they’d be tearing strips off their e-commerce and merchandising directors.

That’s because automated merchandising is already being used by some forward-thinking retailers. They are relying on machine learning to optimise every aspect of the online store - from search to merchandising - to react to trends in real time, to ensure every visitor sees the right products every time, and to do that laborious data work that computers are built for (and humans are not).

By way of example, in an A/B test by Nordic beauty retailer Kicks, a home page was manually produced by merchandisers using their experience and knowledge and pitched against an automated page completely delivered using AI and machine learning.

The results were striking. The automated page produced 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.


That is how you transform your website’s conversion rates - not by just doing more of the same things that are proven to fail. Not surprisingly, the list of retailers employing this kind of technology is growing fast - brands like Fortnum & Mason, and NA-KD to name but three.

If you can combine the great results from traffic generation and the results that AI merchandised websites deliver, it is clear there is an opportunity not only to take the horses to water, but to allow them to drink all they want.

Crossing the online merchandising chasm

That being so, why is it that retail CMOs are not beating down their CEOs’ doors with news of this revolution in online merchandising, and demanding immediate investment in AI?


The short answer is fear. Merchandisers are slow to join the AI revolution because of a mistaken belief that they will lose control, fear that computers cannot be trusted to deal with the nuances of merchandising, or just fear for their jobs.

All are understandable, reasonable and wrong:

  • Nuance: In online retail, nuance means predicting shopper behaviour and reflecting it in second by second merchandising decisions - not in agonising over which tops to place next to those gold trousers. Understanding that online nuance is about crunching data, quickly - a job for which computers and machine learning are ideally adapted.
  • Redundancy: Even with AI doing the heavy lifting, the important, strategic and creative tasks will remain the preserve of merchandisers. Freed from day to day grunt work, merchandisers will be able to focus on higher value activities that drive long term success - and take on frankly much more rewarding roles that reflect their skills.
  • Control: In this AI-paradigm, human control is simply elevated to a higher level. Merchandisers don’t control every decision, but they control the strategies to which they conform - for instance switching between strategies to optimise for conversion or profit depending on overarching business imperatives. During a sale or when clearing inventory is the priority, then optimising for conversion is the way to go, but switching back to a profit strategy takes a matter of seconds.

The bottom line is very simple: Online retail without AI will continue to tread water when it comes to conversion - and the retailers who are first to market with AI-driven merchandising will derive not just transformational results, but lasting advantage.

For those that miss the boat, there may not be a bottom line for long - so have a word with your merchandising team and ask why you have never heard of AI merchandising before it’s too late.