RPA has the experience. Why isn't AI recommending them?

We also tested independent media buying, real estate and home brands, and Super Bowl advertising.

Each question connects to expertise RPA promotes on its website.

Our AI visibility tracking system evaluated 256 responses across these questions between 14 and 20 October 2026.

We measured how frequently RPA appeared, which other agencies were recommended, and which websites and sources AI systems used.

This is a focused assessment of RPA's positioning, not a representative sample of every advertising agency search.

These figures measure how frequently each agency was mentioned in the sampled answers. They are not overall market share or a ranking of agency quality.

More revealing was how dramatically RPA's visibility changed depending on the question.

RPA is already highly visible for full-service advertising in Los Angeles.

But in several categories where it has documented experience, AI rarely recommends it.

A 0% result means RPA wasn't mentioned in the answers recorded for that particular question. It does not mean AI would never recommend the agency in a different search.

The findings give us a practical starting point: understanding why the agency's existing expertise isn't consistently reflected in relevant recommendations.

RPA publishes that statement alongside Liu's comments about intentional, purposeful advertising investment.

Yet when we tested the question Which advertising agencies are best for automotive brands?, RPA appeared in only one of 27 responses.

Here is how the tracked agency groups compared:

We went further than the visibility percentages.

Our system recorded 67 citation events to Dealers United's article comparing automotive digital marketing agencies.

Other commonly retrieved material included specialist automotive marketing comparisons and agency pages discussing dealership marketing, automotive retail advertising and cooperative advertising.

These sources offer clues about what information AI systems use to assemble recommendations.

They do not, by themselves, prove why RPA was omitted. But they identify specific evidence worth investigating.

Its franchise-business expertise page discusses decades of experience working with Honda dealers, La-Z-Boy Furniture Galleries franchisees, ampm franchisees and Farmers Insurance agents.

It also describes the challenge of aligning stakeholders with different priorities around a shared brand and commercial objectives.

That's a highly relevant capability for businesses operating through franchisees, dealers, agents and distributed sales networks.

Yet RPA appeared in none of the 27 answers when we asked AI which agencies specialise in franchises or dealer networks.

So who did AI recommend?

When we examined individual responses, we found companies such as:

These agencies were often recommended because their public positioning addressed specific operating problems.

Those included cooperative advertising, local campaign execution, franchisee alignment, dealer activation and performance measurement across locations.

Importantly, we didn't include all of these companies in our original nine-brand competitor dashboard. We identified them by examining the full AI answers.

That's a useful distinction. A broad competitor dashboard can miss the specialists AI actually recommends for niche questions.

Independence is central to the agency's identity and how it explains its client relationships.

Brett Bender, RPA's President and COO, has also publicly discussed the agency's independence, long-term partnerships and commitment to real business results.

Yet the results of our independent-agency questions were very different.

When we asked for large US agencies that aren't owned by holding companies, RPA appeared in seven of 27 responses.

When we asked for the best independent advertising agencies in the US, RPA appeared in none of the 23 recorded responses.

Same underlying attribute. Different question. Significantly different visibility.

This demonstrates why monitoring only one broad-category question can give an incomplete picture.

It also raises an important question about the information AI relies on to establish category relevance.

We identified third-party agency comparisons and industry publications among the sources consulted for independent-agency recommendations.

Some recognised RPA, while others focused on different agencies.

He goes on to emphasise business results and profit margins.

Yet RPA appeared in none of the 27 answers to our question about independent media planning and buying agencies.

It appeared in only one of 27 answers when we asked about agencies that combine creative and media buying.

What the external sources show

One AI-cited article was Ottaway's comparison of independent media-buying agencies.

The article names Horizon Media, PMG, Tinuiti, and Wpromote as examples.

RPA isn't included.

Our system recorded 37 citation events to that article across the audit.

Other frequently cited resources included agency comparisons from The Manifest and media-buying directories.

This doesn't prove that inclusion in any particular comparison would automatically improve RPA's visibility.

But it gives us concrete third-party sources to investigate.

These differences may be explained by different dates, reporting periods or definitions.

But the relevant website sections don't clearly reconcile them.

This is a practical example of why evidence management matters for GEO (generative engine optimisation).

When a buyer, publication or AI system encounters multiple versions of an important performance claim, the context needs to make those differences understandable.

RPA has people with relevant experience to contribute to those conversations.

The challenge is making those perspectives accessible, attributed, and connected to the agency's wider expertise.

An agency doesn't need executives posting daily about how wonderful the agency is.

It needs its specialists to answer meaningful questions that potential clients are already asking.

These differences reinforce why a continuous measurement programme matters.

An agency may perform reasonably well in one environment but be largely overlooked in another.

However, the fundamental GEO activities overlap.

Clear positioning, strong website evidence, authoritative content, trusted third-party references and accessible expert knowledge remain relevant across platforms.

The differences help you decide which opportunities to investigate and refine.

These are initial research priorities, not a claim that every item represents a confirmed technical defect or that every potential improvement will increase rankings.

The next step is to connect the findings to RPA's commercial priorities and implement them through a structured programme.