It’s never been easier to buy goods online; all it takes is a few taps on an iPhone.
Yet e-commerce sales growth is slowing. Global e-commerce sales are slated to rise 9 percent year over year in 2024, compared to 10 percent in 2023 and 17 percent in the e-commerce golden era of 2021, according to research firm eMarketer. That trend is set to continue: It’s predicted the growth rate could drop to as low as 7.4 percent by 2027.
It’s true that more consumers are shifting their spend back to physical stores. But there’s also ongoing fatigue among shoppers who are tired of sifting through similar products from countless sites with the same grid-like format. What online shoppers continue to crave above all else is a type of personalisation they’re used to on social media, where algorithms tailor the user experience to their individual preferences.
“People want to be catered to,” said Juan Pellerano-Rendón, chief marketing officer at e-commerce logistics start-up Swap. “You want to feel special, that you have this connection to the brand and are being treated as such.”
E-tailers are starting to make changes accordingly, employing AI to do so. In April, eBay introduced a page called “shop the look,” where users can see AI-generated images of shoppable outfits based on their browsing patterns and purchase history. The feature has led to an increase in conversion for fashion items on the site, according to eBay’s chief product officer Eddie Garcia.
But not all companies have over $2 billion on their balance sheets to invest in AI-powered individualised shopping feeds, or newer features with no track record for success like ChatGPT-powered shopping assistants. There’s also the fact that consumers aren’t always receptive to these sorts of changes. In May, Levi’s ended a pilot programme to include AI-generated models on its e-commerce site following overwhelming backlash. In March 2023, Swedish footwear brand Morjas shut down an AI-powered size recommendation tool less than a year post-launch, after customers expressed dissatisfaction with the tool’s recommendation abilities.
The brands that are seeing a lift in their e-commerce businesses today are able to cater to their customers’ shopping habits with online store improvements that fit their budgets. That doesn’t always mean splurging on the latest software.
“Figure out what it is that your customer needs from you before making this massive investment in the next technology,” said Lisa Green, co-founder and chief commercial officer at e-commerce software start-up Daydream, which uses generative AI to improve brands’ search and discovery capabilities. “How do you create that bit of magic for your customer that still feels on brand?”
What Do You Recommend?
Brands without the budgets to develop their own generative AI-powered features can use the data that’s already accessible to them to improve critical selling features like product recommendations.
Womenswear brand Vici, for example, began using Mastercard’s personalisation platform Dynamic Yield to identify low-intent shoppers — those who browse but leave quickly without purchasing — and serve them similar styles at the top of any given product page to keep them engaged. The brand also made its search bar more visible on each product page, allowing shoppers to easily find alternatives to sold-out items. Vici expects these updates to increase conversion this year, said Aimee Dudum Colorado, co-founder and chief merchandising and marketing officer.
“While [some brands] have less money, [they are] more nimble and can react more quickly,” Green said. “Where people see the most success is if they can have a small dedicated team that is reacting to what customers are looking for and trying to put quick tech out there without making a huge investment.”
But the companies with the resources to introduce advanced personalisation features also have to be flexible in updating those tools based on consumer feedback.
Ebay’s “shop the look” feature currently works as a lookbook of outfits based on browsing and purchase history, but users can’t dictate which items are included. To rectify this pain point, the company is planning to give users the ability to generate images of outfits based around a single item they like.
Ideally, any improvements made should be done with the intention of getting consumers more excited about shopping with their brand, Green said.
“It’s not just about that first launch. It’s about quickly learning, hearing from customers and making it better and better,” Garcia said. “You have to move fast in this industry. It’s literally changing on the daily.”
Selling a Lifestyle
As well, when brands listen to what customers they want, they can also stumble onto potential site improvements that don’t require cutting-edge software.
For some customers, it may be as simple as the need to see products contextualised before they’re comfortable making a purchase. Morjas, for example, launched a new blog series called “three ways to style” in February after customers asked the brand for more lifestyle content in its annual survey. Since the introduction of the series, which shows its footwear paired with a variety of outfits, from business attire to occasionwear, the brand has seen a 12 percent increase in conversion, according to Henrik Berg, Morjas’ founder and chief executive.
![A triptych of Morjas' three ways to style series](https://img.businessoffashion.com/resizer/v2/RNAVGINY4NDYTCD5TBQXHYTRVE.jpg?auth=00b95798a15b310179b2b08af2d0446474f3d1ac4a4d7de365a25354a49df31b&width=768&height=339 768w, https://img.businessoffashion.com/resizer/v2/RNAVGINY4NDYTCD5TBQXHYTRVE.jpg?auth=00b95798a15b310179b2b08af2d0446474f3d1ac4a4d7de365a25354a49df31b&width=400&height=177 400w, https://img.businessoffashion.com/resizer/v2/RNAVGINY4NDYTCD5TBQXHYTRVE.jpg?auth=00b95798a15b310179b2b08af2d0446474f3d1ac4a4d7de365a25354a49df31b&width=768&height=339 768w, https://img.businessoffashion.com/resizer/v2/RNAVGINY4NDYTCD5TBQXHYTRVE.jpg?auth=00b95798a15b310179b2b08af2d0446474f3d1ac4a4d7de365a25354a49df31b&width=1024&height=453 1024w, https://img.businessoffashion.com/resizer/v2/RNAVGINY4NDYTCD5TBQXHYTRVE.jpg?auth=00b95798a15b310179b2b08af2d0446474f3d1ac4a4d7de365a25354a49df31b&width=1440&height=637 1440w)
Even selling features that seem complicated, such as sizing, can be tweaked without a heavy lift. Last July, rental platform Nuuly added a filter to its product reviews by height, body type and bra size, which helps users determine the items that are best suited for their shapes in response to feedback from subscribers that they needed more help with sizing. The improvement didn’t require software add-ons.
“People want to hear from real people, they want to know people who are like them, who wore that specific kind of product, and that’s going to give them the most confidence when they go ahead and rent that item,” said Joe Stratter, Nuuly’s executive director of technology.
The stakes for making customers happy is high. With digital ad costs showing no signs of declining, giving customers what they want can help brands retain existing shoppers, who are instrumental to sales and profit growth.
“I’m not spending more on getting a newer audience,” Swap’s Pellerano-Rendón said. “It’s really more so about making the audience that I have feel really special about being a part of our community and our brand.”