From ranking pages to being recommended: what agencies need to rebuild for AI search

Most agencies are still selling AI visibility the way they've always sold SEO: audit the site, fix the technical issues, optimise the pages, report the ranking movement. Our latest research suggests that model doesn't map onto how AI actually decides who to recommend, and agencies that don't adjust are going to keep delivering technically sound work that quietly fails to show up in ChatGPT.

From optimiser to signal builder

For our AI Visibility Report, we audited 10,000 US local businesses through the Insites 360-degree platform, correlating AI outcomes against a wide range of SEO, local SEO, listings and review signals. Technical optimisation still matters, but mostly as a baseline. Once a site is reasonably structured and accessible, further gains come far less from refinement and far more from expanding the volume and coverage of signals associated with the business, the same pattern we first spotted when we looked at what AI actually uses to choose local businesses, and what it ignores.

That's a different job to the one most SEO retainers are built around. Practically, it means shifting agency effort toward:

• Increasing the volume of content associated with a business
• Expanding keyword coverage and service representation
• Strengthening presence across local platforms and directories
• Generating ongoing review activity and engagement

The goal isn't to perfect one asset. It's to make sure the business is widely represented and consistently reinforced across the ecosystem.

Content as coverage, not just quality

One of the clearest patterns in the data is the relationship between content footprint and AI visibility: businesses with more extensive websites, more pages, broader service coverage, more frequent updates, were significantly more likely to be surfaced and recommended.

This lines up with something we're already seeing across Insites partners. Many are using AI and automation to build larger, more comprehensive websites, expand service and location coverage at scale, and regularly publish new or refreshed content. Done well, that directly increases the volume of signal a business generates. The key is that the content still needs to be relevant to real services and user intent, structured clearly and consistently, and actually updated, not just padded out. Content here is less about individual page performance and more about overall coverage and clarity.

Recency and activity as competitive levers

Unlike a lot of traditional SEO signals, recency and ongoing activity showed up as important across nearly every section of the research. Businesses with recently updated websites, active content and ongoing engagement were more likely to be correctly interpreted, described positively, and ultimately recommended.

This is a genuine opportunity for agencies working with smaller clients who can't match a bigger competitor's total signal volume. They can still compete by being more active and more current. That reframes AI visibility work as continuous activity rather than a one-off optimisation project, which is a better fit for retainer-based delivery than for a fixed-scope project anyway.

Local presence and consistency remain foundational

None of this replaces the fundamentals. Google Business Profile completeness, review volume, and consistency across directories continued to show strong relationships with both visibility and recommendation throughout the report. Without this foundation in place, everything else is less likely to translate into actual AI visibility, so it needs to stay a standing part of the service, not a one-time setup task:

• Fully optimised and actively managed Google Business Profiles
• Ongoing review generation and response
• Accurate, consistent listings across key platforms

This lines up exactly with where ChatGPT actually pulls its answers from: the website and directories are the two sources that matter most, so they're the two that can least afford to sit half-finished.

Measurement has to change too

This is the part that will feel most uncomfortable operationally. Traditional SEO reporting tracks rankings, traffic and conversions against specific keywords or pages, cause and effect are traceable. AI outcomes don't work that way. There's no fixed ranking to monitor, outputs vary by query and phrasing, and the factors that drive performance (overall presence, consistency, signal volume) are diffuse and harder to tie to any single action.

That's the same split we've argued for elsewhere: technical AEO audits and ongoing AI visibility tracking are two different jobs, and this report only reinforces why. That means agencies need new reporting language, tracking:

• Whether a business is being surfaced or recommended at all, across a consistent set of representative queries
• How it's being described by AI when it is surfaced
• Changes in overall visibility over time, rather than position for a single term

The mindset shift, and the opportunity in it

Put together, this points to a broader change in what "success" looks like: a business that's widely represented, consistently described, actively maintained, and easy for AI to validate and recommend. Traditional optimisation alone doesn't get a client there. Bigger brands still hold a structural edge in building that picture, but as we've written before, SMBs have real, controllable levers of their own.

That's the challenge. The opportunity sitting alongside it is real too. Visibility built on ongoing signal accumulation doesn't fit a "set and forget" delivery model, build the website, fix the listings, move on. It naturally pushes the client relationship toward something more ongoing and partnership-driven, which is a stronger commercial position for agencies than one-off, low-margin project work, provided the delivery actually holds up to the higher expectation that comes with it. It's the model Insites is built around: one platform that runs the audit, tracks the signals over time, and gives agencies a client-ready report for both halves of the job.

This analysis is drawn from Insites' AI Visibility Report, based on a clean audit of 10,000 US local businesses across ChatGPT and Perplexity. Run a free AI visibility check on any business in seconds.