Small businesses get visible in AI search the same way they earn trust anywhere else: by publishing clear, well-structured content that answer engines can quote, and by being a source those engines already rank. There is no separate switch to flip for ChatGPT, Perplexity or Google's AI Overviews. The fundamentals of good SEO, clean structure, fast pages, authoritative content and schema markup, are what make a page both rankable and quotable. What changes is the emphasis: fewer generic how-to posts, more commercial and hard-to-replicate content, and more work on making your brand known before anyone searches.
No one can promise you a citation in an AI answer. The systems are opaque, they change constantly, and even Google will not tell you exactly why one page is summarised and another skipped. So treat AI search as a shift in emphasis to manage, not a lottery to win. The question is not whether it reshapes your traffic. It is which businesses adjust early and which spend three years pretending nothing is happening.
What is actually changing with AI search?
AI tools do not send traffic the way a page of blue links does. When someone asks ChatGPT for the 'best CRM for a 12-person team', they get an answer inside the interface. No links, no click, and the search often ends there unless they ask for sources.
Google's AI Overviews behave differently. They synthesise several sources and show citations as small favicon links. You can appear in an overview and still lose the click, because the overview answered the question well enough on its own. Independent studies suggest click-through rates fall on queries where an overview appears, though the size of that effect is disputed and varies a lot by query type.
The impact is not spread evenly. Informational queries ('how to descale a coffee machine') are answered in-interface far more often than transactional ones ('buy Miele coffee machine London'). If your SEO strategy rests entirely on top-of-funnel content, you are more exposed than a business ranking for commercial intent.
If your search strategy is built entirely on top-of-funnel content, you are more exposed than a business ranking for commercial intent.
Where do small businesses still win?
AI search is good at synthesising what already exists. It is still weak on originality, local specificity and trust for high-stakes decisions. That leaves three defensible positions for a smaller business.
Original research and data
Models train on published content. If you produce something proprietary, a pricing survey, user-behaviour data, real case-study results, you own something the models can reference but not reproduce. That is why benchmark reports and industry surveys keep appearing: they earn links, they get cited in AI answers, and they give people a reason to click through for the full dataset.
Local intent and service delivery
Overviews still struggle with 'near me' queries and service-area nuance. Search 'best family law solicitor Croydon' and you generally get a local pack and ordinary organic results, because the answer depends on location, availability and personal fit. If you serve a local area, your Google Business Profile and local SEO fundamentals matter more than your blog. The same pattern holds in trades and showroom retail, which is why SEO for a kitchen showroom in 2026 still leans hard on local and commercial pages.
Transactional and high-consideration queries
People use AI search to learn, then use Google to buy. When someone is ready to compare prices, check stock or read reviews, they move into a transactional environment. Rank well for product names, model numbers and 'buy' or 'price' modifiers and you are far less exposed than a business living on 'what is' and 'how to' traffic.
What should you actually do about it?
Most small businesses do not need to panic. They do need to adjust how they think about content and visibility. Four things matter most.
Stop writing for keywords alone
The 'one keyword per page, win position three' era is fading. If your plan is 800 words on 'what is marketing automation', you are mostly producing raw material for the next AI answer. Write things that are genuinely hard to replicate instead: long case studies with real names and numbers, detailed process breakdowns, honest opinions only your business can hold. That still ranks and still earns clicks, because it offers something beyond synthesis.
Optimise to be cited, not just to rank
If your content is going to feed AI answers, at least earn the credit. Overviews pull from pages that already rank well, are clearly structured and use schema. So your traditional fundamentals, title tags, heading hierarchy, fast loads, clean code, now do double duty. This overlap is the whole idea behind generative engine optimisation: the work that helps you rank is largely the work that helps you get quoted. Structured data matters here. FAQ, HowTo and Product schema make your content easier to parse and more likely to be referenced, in an overview or in a ChatGPT or Perplexity answer where your page sits in the retrieval set. None of it guarantees a citation, but it stacks the odds.
Build for brand search, not just discovery
AI search changes discovery. Fewer people will find you through a generic how-to article. But once someone knows your name, they still search for it directly, and branded search converts better and is much harder for an AI answer to intercept. That is an argument for the activities that build awareness outside search: speaking, partnerships, PR, a real presence on LinkedIn, email. If people know your name before they search, you are far less exposed to AI-driven disintermediation.
Put your budget behind commercial content
Educational content is being commoditised. Comparison pages, service pages, case studies and pricing content are not. With a finite budget, spend it closer to the buying decision: 'your service vs competitor' comparisons, landing pages for the service-location pairs that matter, tied together with a sensible internal linking structure, and case studies that include scope, cost and result. These pages are harder to displace with a summary and they convert far better than informational posts. It is the pattern behind our best SEO work: for House of Signs, an SEO and social programme built around commercial search helped drive a 210% increase in revenue and 392% organic lead growth; for Granite Direct, SEO, PPC and CRO together produced a 188% rise in revenue and a 5.3x return on ad spend. Those results came from commercial intent, not from chasing how-to traffic.
Educational content is being commoditised. Comparison pages, service pages and case studies are not.
What not to do
There is already a small industry selling 'AI search optimisation' as a separate service. Most of it is repackaged SEO with a new label. You do not need someone to 'optimise for ChatGPT': clear structure, good content, a fast site and authoritative links are the same mechanics that have always worked. If you are weighing up help, what actually makes an SEO agency good is a better filter than any AI-search buzzword.
Do not abandon content marketing either. The businesses deleting their blogs are overcorrecting. Informational content still ranks, still brings some traffic and still builds topical authority. This is a shift in degree and priority, not a wholesale replacement.
And do not assume you are safe because your traffic has not dropped yet. Overviews are still rolling out unevenly across query types, and tools like Perplexity are growing quickly. The change arrives at different speeds for different businesses, and adjusting early is easier than reacting to a revenue alarm. Expecting results overnight is its own mistake, so it is worth understanding how long SEO takes to show results before you judge any change of approach.
Where should a small business start?
If you are wondering whether this affects you, start with three questions.
- What share of your traffic comes from informational blog posts versus service and product pages? If it is more than 40%, you are more exposed.
- How much of your organic traffic is branded versus non-branded? If it leans heavily non-branded, you rely on discovery that AI search is starting to disrupt.
- Do you have original data, case studies or commercial content that is hard to replicate? If not, now is the time to build it.
We treat this as part of every SEO engagement, not a separate workstream, because it is simply what a sensible search strategy looks like now. If you want a second opinion on where your site sits and whether to rebalance, a free SEO audit is a good place to start, or you can talk to us and we will either help or point you somewhere better.
Michail leads strategy, development, and growth at Molo Agency. He works closely with clients on SEO, web design, automation, and custom software projects, with a strong focus on building systems that generate real business growth.