Christmas shopping – some love it, to others it’s a chore and this year for the first time many of us will outsource the annual task of coming up with gift ideas to artificial intelligence.
While traditional internet search, social media – especially TikTok and Instagram – and simply wandering a local high street will still be the main routes to presents for most this year, about a quarter of people in the UK are already using AI to find the right products, according to PricewaterhouseCoopers.
For brands appealing to younger people, the revolution is well under way: the rival advisory firm KPMG says as many as 30% of shoppers aged 25-34 are using AI to find products, compared with 1% of those aged over 65.
Asking a large language model (LLM) such as ChatGPT or Gemini what you should get your father-in-law – rather than typing “whisky” or “socks” into Google or DuckDuckGo – may seem a small change in habits. However, it marks a sea change for retailers accustomed to paying search engines to promote their listings.
LLMs allow users to ask questions in conversational language, perhaps by speaking into their computer or phone. Instead of just providing a list of links, they offer specific suggestions with the potential for big sales for items that are regularly recommended.
The chatbots produce their responses by scraping the internet and inbuilt datasets for relevant information, with some sources given more trusted status than others.
Companies large and small are scrambling to adapt to this new world where the keywords and advertising deals previously central to web marketing hold less importance than the reviewers’ opinions, accurate availability information and product details read by LLMs such as OpenAI’s ChatGPT, Google’s Gemini and Meta’s Llama.
The shake-up may create an opening for independent businesses to cut through online, but some big brands are concerned they will be lost in a wild west where it is unclear how to reach the consumer. Marketers must now appeal not only to shoppers directly but also to their AI bots.
“Retailers can’t buy their way into the search – they have to earn it,” says Emma Ford, the director of digital transformation for PwC UK. “The experience, expertise, authenticity and trustworthiness [of a brand online] help. Sentiment across the internet is really important.”
Several large UK retailers have told the Guardian they already have teams on the case looking at a wide variety of tactics, from making sure they appear in Reddit forums – a key source for some platforms – to responding to reviews on Google or Trustpilot, and ensuring AI models can access the correct product data.
While some say they are being cautious with resources, amid signs certain individual LLMs could disappear as rapidly as they have sprung up, the belief is that this new way of interacting online is here to stay.
Nickyl Raithatha, the chief executive of the online card and gift seller Moonpig, says AI search’s relevance for companies this year is relatively low but his company is well-prepared for rapid change.
He says Moonpig is ensuring its products are picked up in AI search by using generative engine optimisation (GEO) techniques such as “online content with people discussing the best way to make someone happy on Mothers Day” in its own content or on discussions boards and in YouTube videos. He adds: “There is a growing science around this and we are all learning.”
Ford says businesses are still feeling their way into the nuances of how the technology will find and respond to their online presence. Online reviews, for example, are clearly a factor in AI decision-making, but it is not clear how much importance is placed on particular platforms or how they rank against other factors such as reliable availability data, longevity of a brand or secure payment options.
It may be that suppliers that have been around longer and have a broader profile are foregrounded, but their long history of ups and downs could also play against them.
“I do think AI will change retail for the next 20 years,” says Peter Ruis, the managing director of John Lewis. He contends that established brands such as his will benefit from having a strong reputation with the technology in place to sell online, while shoppers could discover they stock items previously assumed to have been available only at a specialist.
In future, industry watchers believe, ChatGPT, Amazon and Google are likely to try to monetise their AI platforms with some form of paid search or featured ads.
More sophisticated “AI agent” models are also being developed – bots that can autonomously perform complex multistage tasks such as seeking out the best deals, placing orders and organising delivery.
For example, it could be possible for these digital secretaries to negotiate offers tailored to particular customers, such as bundling together a number of furniture purchases from various outlets during a move, which have been customised to fit budget, style and delivery preferences, according to the advisory firm McKinsey.
after newsletter promotion
That could lead retailers into allowing their systems to flex product prices to attract particular searchers.
Organising the return of an unwanted item could also be taken on by AI agents, with one acting on behalf of the shoppers and the other for the retailer.
However, such technology is fraught with potential pitfalls. Retailers will need systems that can cope with a potential flurry of queries and to have clear rules on who might be responsible for glitches such as unwanted purchases made by a bot.
In the US, the online marketplace Etsy was the first to team up with ChatGPT to make it possible to pay for goods via the LLM’s instant checkout service. The e-commerce platform Shopify and the retailers Walmart and Target swiftly followed. While the deals do not appear to prioritise their products in searches, the inclusion of a “buy” button for their goods could put them ahead of the pack.
Anna Bancroft, a partner in PwC’s digital transformation team, points out that under current UK rules it is not possible for an AI bot to make a purchase on behalf of a human, and regulation would need to change for such systems to run without human oversight. She says retailers and shoppers are cautious about giving the robots access to customer data and handling payment.
There are also concerns about agents being susceptible to manipulation, as Microsoft has found in research simulations. Meanwhile, tech retail players are becoming territorial about who gets to crawl whose data.
Last month, Amazon sued the AI company Perplexity over its shopping feature that automates placing orders for users. Amazon accused the startup of covertly accessing customer accounts and disguising AI activity as human browsing. Perplexity hit back, defending users’ right to delegate their shopping to AI agents and calling the suit “a bully tactic to suppress competition”.
In this rapidly shifting landscape, Ford suggests independent retailers may have a chance to shine. “Independents have potential to go faster,” she says, with the ability to respond nimbly without having to sign off large budgets.
Michelle Ovens, the founder of Small Business Britain, which advises independent retailers on how to survive on the changing high street, agrees. “[Independent businesses] don’t necessarily need to spend a lot of money. You don’t necessarily need a big team,” she says.
Ovens advises local shopkeepers to ask AI platforms themselves how best to make sure they can appear. “Be clear about who you are,” she says, with a description making clear that you are an independent specialist, up-to-date pictures and “encourage customers who have got experience of the brand to give a good review”.
However, all of this should not stand ahead of making a website engaging and easy to shop on, Ovens adds. “There will not be a dramatic shift this Christmas. We’ll see change over time and operators will rise to the challenge.”