Win AI Search in 2027: 3-Stage System
57.51% of US adults have already bought a product recommended by an AI chatbot, according to a Semrush study of 2,300 people.
More interestingly, 57% have decided not to buy something because a chatbot told them not to.
AI tools are already deciding whether customers discover your brand, choose you over a competitor, or walk away. To win AI Search going into 2027, a decent visibility score isn’t enough. You need a repeatable system for finding your gaps, building the signals AI tools look for, and improving continuously.
Tim Cameron-Kitchen, Founder of Exposure Ninja, walks through that system using real results from businesses Exposure Ninja works with, plus a live demo in Semrush One.
What AI Search Domination Actually Looks Like
Tim Cameron-Kitchen is clear about the target:
“Today our goal isn’t just, oh, let’s improve our visibility in AI. We want AI domination. We don’t just want one of our pages to be cited in an answer. We don’t just want a high AI visibility score. We want to consistently appear across multiple AI platforms for the sorts of questions and queries that turn into real money for your business.”
That means showing up across Gemini, ChatGPT, AI Mode, Perplexity, and Claude, for commercial queries. As Tim Cameron-Kitchen puts it, “we want the questions that imply somebody is searching for a product or a service to buy from.”
Elite Renewables: Doubling Leads While Cutting Ad Spend
Elite Renewables is a heat pump, solar and battery installer owned by Exposure Ninja. Each customer can be worth anywhere from £10,000 to £300,000.
Search for “best heat pump installers” and Elite Renewables consistently takes the top two organic results and the top two AI Overview results. That visibility carries through into Perplexity and Google’s AI Mode, too.
Over the last eight months, Elite Renewables’ lead volume has roughly doubled, while paid ad spend has been cut significantly.
An eCommerce Brand Growing AI Revenue by 275%
An eCommerce brand Exposure Ninja works with generated 5.3 million Google search impressions for 47,000 clicks. Its AI Overview impressions are now a quarter of that figure, and climbing fast.
Traffic from AI platforms, led by ChatGPT, now tops 1,000 visits a month, with revenue from AI platforms up 275% year on year.
“We don’t just want to be seen in the answer. We actually want to be the one that’s driving the clicks or driving people to go and search for your brand on Google. We want that to turn into leads, sales, and website traffic,” explains Tim Cameron-Kitchen.
Stage 1: Find the Gaps in Your AI Search Visibility
Before you can improve anything, you need to know where you stand. Tim Cameron-Kitchen explains how the gap is defined:
“We need to know where you’re visible at the moment and where you want to be visible. That is the gap. We can then work on filling the gap by giving these AI tools what they need to see in order to start recommending you in those opportunities where currently your competitors are being put forward.”
You could run hundreds of manual searches, or you could use an AI visibility tool built for the job, like Semrush One, which tracks visibility across Google AI Overviews, AI Mode, ChatGPT, Perplexity and Gemini.
In the demo, Tim Cameron-Kitchen analyses the UK law firm, Irwin Mitchell. It has an AI visibility score of 49 out of 100, which is respectable. But three metrics tell the real story:
- Mentions are how often the brand appears in AI answers.
- Citations are how often a page on its website is referenced.
- Cited pages are how many individual pages are being drawn from.
For Irwin Mitchell, all three have been sliding since early 2026. That’s the cue to dig into who’s winning instead.
Check Which Topics Your Competitors Own
The competitor research section shows topics where competitors are mentioned in AI answers and you aren’t. Not all of them will matter. One law firm competitor had visibility for crossword solving, thanks to AI confusing it with a different website.
So Tim Cameron-Kitchen’s advice is that “before you start flogging yourself for doing such terrible work with your AI Search Optimisation“, check you actually want to win those races.
The real gaps are commercially valuable. Irwin Mitchell has no AI visibility for cohabitation law, workplace sexual harassment, or prenuptial agreements in the UK, even though competitor Slater and Gordon appears for the latter, and a decent share of that search volume is commercial.
“If these are areas of law that Irwin Mitchell wanted to help with and they wanted to get more clients for, they could quite simply build out pages on their website and build out some more content on their site about these topics to increase their chances of being cited by the AI tools in these areas,” explains Tim Cameron-Kitchen.
Work through this, and you build a shopping list of content to create or strengthen.
You can test this process with a free 14-day trial of Semrush One at exposureninja.com/semrush-one*.
Stage 2: Build the Proof AI Tools Need to Recommend You
Once you know the gaps, you need to fill them with evidence.
“If we’re going to get these AI tools to recommend us, we need to give them the evidence in order to recommend us,” Tim Cameron-Kitchen, Founder of Exposure Ninja
Strengthen Your On-Site Signals
Depending on why you’re losing to competitors, that evidence could include:
- Technical SEO, so search engines can crawl and understand your key pages.
- Clear brand positioning, like HubSpot’s concise “What is HubSpot?” section that AI tools can repeat to users.
- Niche content, such as Mortgage Genie’s page on mortgage options after one year of self-employment, which targets a surprisingly competitive term.
- Clear page structure, with summaries, comparison tables, FAQs and schema.
- First-party data that AI tools can cite.
- Firsthand experience from your team.
That last point matters because, as Tim Cameron-Kitchen notes, AI tools “don’t have experience of anything. They are word prediction tools.” Your experience is something they can’t replicate, so they’ll pull it from your site.
Comparison tables are especially useful. They tell AI tools when to recommend you over a competitor. Keep them honest, but frame them around the customers you genuinely serve best.
Get the Rest of the Internet Talking About You
Your own website is only part of the picture. Muck Rack’s December 2025 analysis found that 94% of generative AI citations came from non-paid, non-brand-owned sources, like news sites, blogs and Reddit.
That’s where digital PR comes in: finding the publications AI answers draw from and getting your brand featured in them. Tactics include data-led stories, reactive PR, creative campaigns, product PR and resource-led PR.
Press releases sit firmly at the bottom of that list, because, in Tim Cameron-Kitchen’s words, “nobody needs another press release.”
For products, review roundups on magazine sites feed directly into AI Overviews, so getting featured there puts you in front of buyers who trust the AI’s top pick.
Decide What You Want AI to Say
AI tools describe your brand based on how it’s consistently described online. If you want a phone case known as lightweight, that word needs to appear in your comparisons, your briefs to reviewers and your press coverage.
“ChatGPT isn’t weighing this phone case; it’s just seeing, ‘Oh, this phone case, often mentioned as lightweight. Therefore, I’m going to present this phone case as lightweight,’” explains Tim Cameron-Kitchen.
Many brands skip this step, and some discover internal disagreements about their positioning when they finally sit down to discuss it.
Work out how you actually want to be positioned in the AI Search results. Then, when we’re going out to the press, when we’re publishing content online, we can make sure we’re tying it all in with that positioning.
Tim Cameron-Kitchen
Founder
Stage 3: Execute at Scale With a Repeatable System
AI Search visibility isn’t a one-off project.
“This isn’t something that you play with on Sunday afternoon, by Monday you’re ranking and you never have to worry about AI visibility again. This is something that’s constantly changing, and you need to be feeding the beast all the time, giving these AI tools more and more information to justify them recommending you,” Tim Cameron-Kitchen, Founder of Exposure Ninja
How Opal Turns Data Into Action
That’s why Exposure Ninja built Opal, a framework with four stages: Observe, Plan, Act and Learn.
This summer’s heatwave pushed UK interest in air conditioning to record highs. Opal’s Observe stage, which monitors analytics, search volume and Search Console data, spotted the spike for Elite Renewables.
Opal and the Exposure Ninja team then planned and produced a large volume of air conditioning content, targeted at different customer types and optimised for AI Search. That included data-led pieces built for press coverage.
Visibility and enquiries for air conditioning both rose. Opal then turned what worked into a playbook for next time.
Don’t Ignore Paid AI Search
OpenAI began rolling out ChatGPT ads in the U.S. in February 2026 and, according to OpenAI, reached a $1 billion annualised run rate in under 200 days. Ads launched in the UK in August.
“You can’t pay to influence ChatGPT’s organic recommendations, but you can pay to have your ads shown next to those organic recommendations,” says Tim Cameron-Kitchen.
For one UK B2B client, the first six weeks delivered 29,000 impressions and 330 clicks at an average of £1 per click.
It’s a small data set, but the differences from Google Ads are already clear. There are no keywords: you describe your audience, and the AI handles targeting. Conversation volume drives reach, with one campaign generating 10 times the impressions of a near-identical one.
Clicks aren’t the same as conversions. Of the two ads tested, Tim Cameron-Kitchen says, “the much more exciting one generated loads and loads of traffic. The boring one generated all of the leads.”
ChatGPT ads are still evolving quickly, so check current platform rules before you launch.
If you want to win AI Search in 2027, Exposure Ninja’s expert team can find your gaps, build the proof, and run the system that keeps AI tools recommending you. Book a consultation call to see what that looks like for your business.
*Disclaimer: Exposure Ninja may get a commission through the marked links above, at no cost to you
