A good deal happens between a question and a purchase. People still search the way they always have, but now an AI-generated result sits at the very top of the page, pulling the best options into one neat response. Increasingly, users are bypassing the search engine entirely and heading straight to social platforms like Instagram and TikTok to find reviews from real people.
This is answer-engine optimisation. An AI assistant curates its response by scraping the web and choosing what it can grasp quickly and trust to be accurate. More buyers are discovering products and services this way every month. Having content that is clear and relevant is what increases the chances of AI presenting it.
Winning the recommendation
OpenAI counts around 900 million people using ChatGPT every week, with roughly 50 million shopping questions put to it every day, and research from fintech platform Adyen shows the number of shoppers turning to AI for help has more than doubled in the past year. So, how does a product or service become recommended?
Not only are more people landing on websites through AI results, but they may even be better customers. Research from Adobe in 2026 found that shoppers arriving through AI converted 42% better than those from other channels. Adobe also found revenue per visit from AI traffic was 37% higher than non-AI traffic, with engagement 12% higher, time on site 48% longer and 13% more pages per visit.
For years, marketing was built around winning ‘the click’ whether that be through a search engine, or an ad. Once someone landed on a brand’s website, it had a chance to win their business. Now however, AI assistants tend to answer questions directly, meaning the click may never happen. But if the assistant recommends a brand, that name sticks, so the aforementioned can still win the customer. A recommendation works because it feel like it comes from a source the buyer trusts.
In many cases, the recommendation matters more than the click ever did. People are more likely to trust a suggestion that feels honest, relevant and useful. Instead of “selling” to customers, the goal is to become the answer they are looking for.
Earning a place
Search engines rewarded visibility, and for years businesses invested heavily in improving their rankings. AI assistants work differently. They draw information from many sources, compare what they find and favour brands whose information is clear, consistent and easy to verify.
A brand earns its place in these answers by being its own clearest source, by stating in plain language what it sells, what it costs and who it suits. The assistant is, in effect, taking a brand at its word, so the words need to be clear and true. An assistant does not see a brand the way a customer does, through a logo or a feeling, but through the substance of what it can find and confirm. To the machine, a brand is the sum of what is clearly said about it, on its own pages and elsewhere.
Search rankings still matter, but not in quite the same way. AI assistants build their answers in real time, drawing on the information available to them. They don’t necessarily favour the ones with the highest search rankings, which creates an opportunity for newer and smaller businesses. A young company with a clear product and useful information can appear alongside much larger competitors, simply by answering the questions people are asking.
As always, reputation matters too. A company’s website is no longer the only place AI learns about it. Assistants gather information from reviews, news articles, forums, social media, creator content and third-party websites. In some cases, a brand’s own site makes up only a small share of the sources used to generate an answer.
A brand should explain its products and services the way it would to a customer, clearly, honestly and without unnecessary jargon. It means answering common questions, keeping information up to date and making the basics easy to find.
A Search for Success
Success looks a little different. It is no longer just about where a brand ranks in search results, but whether it is recommended, described accurately and trusted enough to become part of the answer. Visibility still matters, but credibility matters more.
Perhaps the most encouraging part is that this is open to businesses of every size. A place in an AI assistant’s answer cannot simply be bought, it is earned over time by being useful, trustworthy and easy to understand.
More and more buying decisions are now taking shape inside conversations between people and AI assistants. Brands that fail to appear risk missing an important moment in the customer journey, while those that do appear are introduced to customers by a source they already trust, often just as they are ready to make a decision.
AI assistants tend to favour the same things people do, such as clear information, honest communication and businesses that are genuinely helpful. In many cases, improving how a brand is understood by AI also improves how it is understood by customers.