Over the past few years, GenAI transformed from being a buzzword to being adopted
by many of us. But that adoption is in most cases limited to GenAI simply being our own
personal in-browser assistant. Combining GenAI with the physical world is still largely
undiscovered territory – especially in retail. And that’s a shame, because the opportunities
emerging are very exciting! This article explores some possibilities of how GenAI can be
applied to stores. For the sake of inspiration and preparation, because we need to prepare
today so we can act tomorrow.
Our shopping streets have changed, but not that significantly.
Over the past years, many retail thinkers claimed that our shopping streets will change. And some of these prophecies were right:
• Better integration between online and offline shopping (driving conversion)
• Contactless payments, or even cashierless stores (bringing operational efficiency)
• Experience-driven and Instagrammable store designs (boosting shopping experience)
To be fair: the above changes are major improvements to the stores of ten years ago. But they’re not the paradigm shift our shopping streets have been waiting for. They are incremental changes, optimizations even, to traditional retail. And now that GenAI is available to all of us – our shopping streets are on the brink of a transformation more radical than ever before.
GenAI brings a paradigm shift to our shopping streets
Generative AI interprets and processes unstructured data and generates new and original outcomes.
Applying that to our shopping streets gives us opportunities for (at least) product advice, try-ons,
store windows and store improvements.
Next generation product-advice
Store employees provide advice from their own knowledge based on the input that we provide them. This creates a dependency on two variables: 1) input: the information that we give or do not give and 2) output: the knowledge and experience of that specific store employee.
Fast-forward to the future: let’s say that instead of a store employee, we have an interactive voice-controlled virtual assistant fueled by GenAI. Perhaps an Expert Emma, Fashionable Frank, or Wise Wendy. This changes the input and output significantly:
Virtual try-ons
Think about the things you have to do to try-on a piece of clothing: walk around the store, stumble across an item you may like, select a color, assess what size you need, check if that size is available, assess if the price is right for your budget, take it to the fitting room, decide on the purchase. And repeat this process if the outcome is suboptimal. That’s a whole lot of process steps! And each of these steps provides the opportunity to not make the purchase. A waste of the customer’s time, and a risk for the retailer of not making the sale.
Fast forward to the future: imagine GenAI-powered mirrors in the store, seeing you standing before the mirror, recognizing you, knowing your style, connecting to your previous purchases and estimating your size. This mirror then acts as a personal shopper suggesting spot-on fashion pieces and instantly having you try them on virtually!
This reduces every risk of not making the sale for the retailer:
• The entire assortment is factored in, in every color and every variant
• You may virtually try on pieces that you would have walked past before
• Available sizing is irrelevant, because if it’s not available you can have it ordered instantly
• No more two sizes to the fitting room; the mirror instantly recommends one based on its estimation
• The price becomes less relevant once you’ve seen it on you and have fallen in love with it
• Happy customer because the purchased item is actually relevant, hence reducing returns
Custom store windows
Store windows have one purpose: driving traffic. So, the content needs to be spot-on relevant for the customer. Fast forward to the future: what if each store window would be suited exactly to the individual(s) standing in front of it? One or multiple screens playfully arranged, recognizing the characteristics of who is walking in front of the store: is it an individual, a couple, a group of friends or a family? What are the genders, sizes, ages and styles of the people standing in front of the window?
Based on the information interpreted, the store window presents the right assortment to the persons in front of it. And doing this right – what if it could also lure you in by dynamically offering the right kind of promotion or discount that you are likely most susceptible to?
Continuous improvement in store operations
Changes in store layout are often driven from headquarters – more often than not in a one-size-fits-all approach. But stores aren’t one-size-fits-all (for many reasons), making every layout change suboptimal at best. Fast forward to the future: imagine tactically placed cameras and sensors in the store, continuously analyzing customer behavior and movement throughout the day. And then imagine these images connected to a GenAI engine that recognizes patterns.
• What areas of the store are busier than others, at what moments?
• Where do customers seem lost, or walk inefficiently?
• Where is the customer’s interest gained, and where is it lost?
• Where do bottlenecks emerge and why?
• How effective are store employees to support the customers?
And then imagine the GenAI engine self-recommending alterations and initiatives to optimize accordingly? Instant optimization, custom for each store!
Prepare today, so you can act tomorrow
The thing with futurizing is that it often feels like it’s still far away. That’s not the case with the examples described here. Early adopters in retail are moving in the GenAI direction as we speak, seeking out new ways to stay relevant for their customers. We know this, as IG&H is already supporting first movers in this topic. The secret of these early adopters? Doing. Sometimes it’s that simple. Tomorrow’s retailing has always been defined by preparations taken today, and that’s still true for the future of retailing.
Start your GenAI journey in 7 steps:
Think MVP: create an initial scope with basic features that helps build excitement
Get a team going: a handful of dedicated forces with knowledge of systems and data
Set a deadline: be bold and strive for initial results in 4 – 8 weeks
Go low-code: build the MVP app on a low-code platform (such as the Microsoft Power Platform)
Isolate: minimize dependencies on other systems by isolating (data)environments
Test: experiment, test, evaluate and update continuously
Expand: integrate with live systems, improve and prepare for roll-out
Cazijn Langeler
Director Retail
+31610545518