Let's say you just bought a suitcase for your upcoming holiday. Over the next week, you will be given recommendations on different suitcases. However, you have your suitcase. You don’t need another (unless you have an excellent luggage add-on package); at this point, product recommendations are just a nuisance.
If you’ve had this experience, you’re not alone.
According to a recent Statista report, 43% of US respondents struggled with being marketed products they had already bought. Anecdotally, there is no doubt that this is a global problem. It's not only a wasted opportunity to showcase a more expansive display of products; it’s a significant turn-off for consumers and their opinion of your brand.
A new collection of consumer data
Thankfully, generative AI has the potential to make this a thing of the past. The fashion industry has confined fashion personalisation to marketing recommendations for customer sub-segments based on past purchases or online browsing history, held back by talent and technology constraints.
With generative AI, there is scope to further this by capturing a more comprehensive range of data to understand the entire customer journey, not just from the last week but from the previous two to three years. As a result, brands can get to know their customers' preferences and shopping habits far better too.
Enabling Deep(er) Learning
This data and understanding can also be used to generate content. To return to the holiday example, rather than suitcases, you might be recommended sun cream, a beach bag, a hat or sunglasses. But it runs deeper than that. With the ability to assign specific descriptors to products, retailers can know if Piz Buin is more your style than Supergoop! based on this better representation - or holistic view - of the buying journey.
This is because generative AI can understand many different similarities products have, such as an oversized silhouette or pearl buttons. It literally learns an individual’s style to optimise their shopping experience. Essentially, it gives every customer their own personal shopper.
The discovery dilemma
AI is also enabling brands to solve the discovery dilemma. Consider this: most retailers have vast amounts of pages on their website, and consumers only click through the first few. AI will help retailers surface the most relevant things - the products customers have an affinity for - and use the information to put them on the first few pages they see. This is increasingly important for consumers. For example, 75% leave a site after two minutes if they cannot find the desired product or service.
Representation extends beyond consumer personas
Generative AI’s ability for personalisation and inclusivity extends beyond recommendations. The technology has reached a new frontier in generating hyper-realistic images. An excellent example of this is Levi’s announcement that it will be building fully customised fashion models through advanced generative AI.
The aim is “to try to create a world where every customer that comes to shop at Levi's can see our products on a model that looks and feels like them” - covering every combination of race, ethnicity, age, body size and type. In the future, personalisation will be taken a step further, with customers inserting themselves into product photos on e-commerce sites.
Transparency will be crucial for success
It would be remiss not to add that Levi’s AI-generated models have drawn criticism for failing to advance real diversity. One article in The Cut headlined its story, “Levi’s you can just… hire diverse models?” Success will rely on all assets being reviewed for accuracy and bias. However, if brands can get it right, the results could prove a phenomenal success.
Straight-forward personalisation of customer experiences is no longer good enough. Consumers increasingly expect brands to know them better than they know themselves. Brands must meet these challenges with bespoke, hyper-personalised experiences for each consumer - and generative AI can make this happen.
