Brands combining SEO and PR have always outperformed brands relying on just one. New 2026 research strengthens the case, especially for AI search visibility.
Almost half of the brands ranking on Google’s first page are invisible to ChatGPT for the same queries. That gap has a structural cause and a familiar fix. Backlinks and referring domains have been measurable evidence of integration paying off for two decades. AI citations are the new layer, sitting alongside the long-running link signals as evidence the work pays. Both functions still work on their own, but they’ve always worked better together. AI search has made the combination matter more.
Google rankings no longer guarantee AI citations
EMGI’s SaaS AI Citation Gap Report, published April 2026, analysed 150 SaaS companies across 120 keywords. 44% of brands ranking in Google’s top 10 received zero ChatGPT citations for the same queries. Conversely, 81% of ChatGPT brand mentions came from companies outside Google’s top 10.
The reason is structural. Google works from an index of pages it has crawled and stored. AI search systems pull sources when a user asks a question, then synthesise an answer from the places they trust. The two systems reward different signal mixes and brands captured in both are those running both functions.
Your website is still the foundation
Your website must do its job. If it doesn’t clearly communicate who you are, what you do, and how well you do it, the rest of the visibility stack has less to work with. Owned content is a baseline. AI models retrieve and read what’s there and so it follows that a thin website limits the impact of off-site efforts.
Where AI citations actually come from
The sources of AI citations beyond the website will feel familiar to digital PR teams: mentions, backlinks from credible publications, branded search demand and presence in trusted channels.
Airops’ 2026 State of AI Search found that brands investing in off-site presence are 6.5× more likely to earn AI visibility with around 85% of those mentions living on external domains.
LinkedIn is worth singling out. SEMrush’s analysis of 325,000 prompts found LinkedIn is now the second most-cited source in AI search behind Reddit, with around 11% of answers referencing a LinkedIn URL. ChatGPT and Google AI Mode pull most LinkedIn citations from individual posts rather than company pages, making employee thought-leadership the lever for B2B brands. We treat that as part of brand footprint anyway and the data confirms it.
Where the combination outperforms
The compound effect is critical. Brands that earn a Google ranking alongside an AI citation are 40% more likely to resurface in subsequent AI runs than brands cited only once. A B2B SaaS ranking #1 on Google can be invisible in ChatGPT for the same query if competitors have built more mention volume on review sites and trade press. The ranking still earns the click in traditional search. It just earns less from the AI answer.
The dual-signal payoff also lifts traditional performance. Seer Interactive’s analysis of 25.1 million organic impressions across 42 organisations found that brands cited in AI Overviews earned 35% more organic clicks and 91% more paid clicks than uncited competitors. Earned coverage doesn’t only feed AI search – it tightens performance across the whole search ecosystem.
We see the engagement signal in our own client data. For a UK housebuilder, ChatGPT referrals delivered 3.80 pages per session between October 2025 and February 2026, ahead of every other traffic source on the site (Bing 3.00, Google 2.77, direct 2.65, Rightmove 2.24, Facebook 1.67). Volume was modest at 230 sessions over five months, but ChatGPT was the clear leader among AI sources, well ahead of Gemini (13 sessions), Copilot (7) and Perplexity (3). These visitors viewed more pages per session, aligning with broader findings about AI-referred intent.
AI citations are the latest measurable output of the same work that has earned the third-party coverage backlinks depend on. Brands consistently visible across Google and AI search are running them as one programme.
AI models are getting more selective
ChatGPT cites only 15% of the pages it retrieves. When GPT-5.3 Instant became the default earlier this year, it cited 20% fewer domains than its predecessor and subsequent updates have continued the trend. Brands that maintain visibility through those changes tend to run a steady earned-media programme. Whether any specific citation today survives the next model is uncertain – the programme that produces the citations is more stable.
Three pieces of an integrated programme
For brands already running SEO and PR as one programme, this argument is settled. For everyone else, the obstacles are familiar: separate teams, budgets, and reporting lines. The case for integration is now stronger.
The work splits into three pieces.
This isn’t about rebuilding marketing around AI search. AI search is still a small share of total traffic for most brands. The real value is that the same work improves both AI and traditional search performance, positioning brands ahead of future shifts.
What this changes in the planning brief
The argument for integrating SEO and PR isn’t new and the evidence isn’t either. What’s new is the type of evidence. AI citations have joined the long list of signals that earned coverage produces. For integrated teams, the data validates the model. For everyone else, it makes the business case for integration harder to ignore.
GO BIG
OR go home
