AI SEO agency: what changes in scope and reporting
What an AI SEO agency should add to a classic SEO scope, what stays the same, and how reporting on AI answers differs from rankings and clicks.
By Roozbeh Nazari · CEO
"AI SEO agency" has become a common label, and a search for "ai seo agency" turns up providers that mean very different things by it. For some providers it means using AI tools to produce content faster. For others it means work aimed at visibility in AI answers, such as Google's AI Overviews and AI Mode or assistants like ChatGPT and Perplexity. Before you compare proposals, it helps to know which parts of the scope really change when AI search is in the brief, which parts stay exactly the same, and what honest reporting on AI visibility can and cannot show.
What stays the same in AI SEO services
Much of the foundation does not change. Google states that there are no additional requirements to appear in AI Overviews or AI Mode, and no special optimisations are necessary beyond sound SEO practice. Pages still need to be crawlable and indexable, content still needs to be helpful and clearly written, and internal links still tell search engines which pages matter. Google also notes that you do not need new machine-readable files or special markup to appear in these features.
That has a practical consequence for any AI SEO services proposal: if the technical foundation, content quality or site structure is weak, those fixes come first. An agency that jumps straight to "AI optimisation" without checking indexing, page experience and content quality is skipping the part that most of the outcome depends on.
What changes in the scope
Some work is new, or at least much more important than before:
- Question and topic mapping: listing the questions your buyers ask in conversational form, not only the short keywords they type, and checking which of them trigger AI answers.
- Answer-ready content: clear definitions, direct answers near the top of a section, comparison tables and specific, well-sourced facts that an AI system can summarise accurately.
- Entity and brand consistency: making sure your company name, services, locations and key facts are described the same way on your site and on the profiles and directories you control.
- Source monitoring: tracking which pages and third-party sources AI answers cite for your topics, so you can see where your content is missing or misrepresented.
- Crawler access decisions: reviewing robots.txt and snippet controls for search and AI crawlers, so that access is a deliberate business decision rather than an accident.
None of these tasks requires a separate website or a separate content strategy. They are extensions of good SEO, and a well-run agency folds them into the same roadmap. You can see how we structure this work on our AI search visibility page.
How reporting changes
Classic SEO reporting rests on rankings, impressions and clicks. AI search blurs these. In Google Search Console, traffic from AI features is included in the overall Performance report under the Web search type, not in a separate AI report. Assistants outside Google have no equivalent of Search Console at all. This means an AI SEO agency has to combine several imperfect signals rather than rely on one dashboard.
A sensible report usually includes:
- A fixed prompt set: a documented list of questions, run on a regular schedule in the assistants that matter to you, recording whether your brand is mentioned or cited and which competitors appear.
- Citation share for key topics: how often your pages are among the cited sources, tracked over time rather than as a single snapshot.
- Search Console trends for the pages you expect AI features to draw on, with the caveat that AI traffic is mixed into normal web results.
- Referral traffic from AI assistants in your analytics, labelled as a partial view because not every visit carries a referrer.
- Business outcomes: leads, enquiries or sales from these pages, which remain the numbers that justify the work.
AI answers vary from one run to the next, so single screenshots prove little. Ask for the method, the prompt list and the dates, so that the numbers can be repeated. Our guide to measuring visibility at prompt level explains one way to set this up.
Red flags in AI SEO proposals
Be careful with proposals that promise a guaranteed mention or citation in ChatGPT, AI Overviews or any other system. No agency controls how these systems choose sources, and Google's general guidance on hiring an SEO already warns against guarantees. Also watch for:
- Large volumes of AI-generated pages sold as "AI SEO". Google's spam policies treat mass-produced pages with little value as scaled content abuse, whatever tool produced them.
- Special files or markup presented as a shortcut to AI visibility, when the search engine's own documentation says none is required.
- Reports built only on screenshots, with no fixed prompt set or repeatable method.
- A scope that has no line for technical SEO or content quality at all.
A good AI SEO agency explains what it can influence, what it can only monitor and how both connect to your business goals. If you want to compare approaches, start with our services overview.