How will AI agents change the way we shop? FutureWeek caught up with Ben Wiener, Global Head of Cognizant Moment, Cognizant’s digital experience business, to explore what the company’s recent report says about changing customer expectations because of AI, what brands need to do to be picked up by AI agents, and more.
Cognizant recently released a new report, what were the findings of this report?
In New Minds, New Markets, we identify consumer expectations around AI in their shopping. We found there isn’t anywhere near as much innovation, maturity or experimentation as customers are expecting. That’s the biggest threat and opportunity for major companies is how they close that gap between customer expectation and what they could deliver. Legacy brands are at risk of being disrupted by direct to consumer brands built on AI models as opposed to digital models that will steal market share if they don’t keep up.
How will the way consumers interact with AI tools change over time?
There’s a general expectation from consumers that AI works like search, so anything you search in Google you also should automatically search in AI. That’s actually bad for the planet – to treat AI search the same as traditional search. We want consumers to use AI for where it adds value and not as a substitute for sifting and filtering through content.
Consumers expect complicated query-answers to come to them as fast as traditional search, and this isn’t realistic.
What are brands doing to respond to consumer use of AI tools?
From a CMO perspective, there’s a whole other level of data fluency and technical fluency that’s going to be required to take advantage of an AI and agentic AI era.
In the early days of Google, there was the continuous gaming of the Google algorithm to appear in search results as best as you could. We don’t know the exact details of how AI models are discovering brands online – Microsoft may do it in its own way, while Gemini might do it in another way. We’re not at a point where one AI model has market share that that’s where you focus.
As brands build their own agents, those will work in a different way too. So far, it’s been about being omnichannel. Now, it’s an ‘omnibuyer’ approach where we have to appeal to numerous entities – whether that’s a human, an agent that works on behalf of a human, or an AI built into an existing platform.
What would you recommend as a starting point for brands to connect with customers using AI?
Your inventory and what you sell becomes increasingly commoditised in an AI age, so you need to have a sharp definition of what your brand stands for, not what your brand makes, in the age of AI. Increasingly, agents will act for consumers through focusing on their values. What are the attributes of a brand beyond the functional that creates that brand differentiation? Is your brand green? Is your brand Fair Trade?
People will no longer specify exactly what they want, they will rely on suggestions from their own AI agent. Think of a time when a consumer will be able to grant its shopping agent access to its Instagram account so it knows what clothes it wears – so it’s shopping based on style, not based on preference for a specific brand.
What about those brands that aren’t strong in their niche?
The other side of it is ensuring availability of real-time data and inventory for people who are more convenience-minded. Brands who are feeding their inventory, pricing, and delivery times in real-time into AI have an advantage because they can provide instant gratification. For these consumers, they’ll never have to look at something that isn’t in stock, or isn’t available in their size, or isn’t available in a specific colour.
It’s also about those brands that reduce the cognitive load for consumers and align to what they’re prioritising – so these brands have a stake too.
Does this broaden or diminish brand audiences?
I think it broadens audiences for brands, because if a trusted AI is telling a consumer a brand is worthy of consideration, then a consumer will listen to that agent even if it’s a brand they’ve never heard of.
Will customers have fewer touch-points with brands?
We’re moving towards a post-click, post-wireframe world where interfaces will be conversational. This doesn’t mean, however, that brands won’t have an opportunity to create experiences for their customers.
For example, if I’m Abercrombie, there’s a very distinctive look, smell, and feeling to an in-person store. A consumer might not navigate to a website in the same way, but might be shown a video of the product they’re looking for. There are ways of creating a brand experience in AI that goes beyond a flat e-commerce shopping experience.
With the rise of AI agents, will brands have to think about what character traits represent them fully through their agent?
There are some considerations that need to be made around how much autonomy is granted to brand agents.
Does the rise of shopping agents get rid of the randomness in purchasing decisions?
What are all the different roles that shopping plays in our lives? We purchase practically and celebratory things. AI might know that every Wednesday we get a pizza, and it might prompt us to order the pizza on the way home from picking up the kids. It might give you a cool new suggestion for a new restaurant you would want to try. There’s still an element of discovery in there, as opposed to proactively streamlining and automating tasks that are repetitive. But there’s no doubt, AI has the ability to influence our purchasing decisions.
Because of its conversational nature, it seems like we trust AI quite a bit. Are there any dangers to this?
Who owns the AI models we’re taking purchasing advice from? If it’s Big Tech companies like Microsoft and Google, they’ll likely prioritise brands they’re aligned with or are partners. Who builds the AI model, controls it, and who pays for it is all a part of the equation. If you can’t spot the product, it’s you.
So, if there are dangers involved in ‘Big Tech’ having a monopoly on these tools, what’s the solution?
In the early days of the internet it was the Wild West. There was no consistency around advertising formats, around payment, privacy, ethics, or copyright. It took a while for consumers to trust using the internet. For AI, we’re going to need industry standards. We know when a Google search result is sponsored, but will we know when an AI chatbot result is sponsored? How we bring people together to agree on a certain set of standards will be key. I think the faster we can create a set of standards that can lead to trust, the faster the overall adoption of AI will happen.
