Berlin, GERMANY, Nov. 11, 2026 (GLOBE NEWSWIRE) -- Peec AI, the AI search platform that helps marketing teams understand and improve how their brands are discovered, represented, and recommended, today launches Primary Sources, a quarterly research series tracking how AI search is changing and how marketing teams are keeping up.
The first edition combines a survey of 700 marketing decision-makers at companies with 100+ employees in the US, UK, and Germany with Peec AI's own platform data on what AI platforms retrieve and cite. It finds that AI search has become a top marketing priority and budget line before it has become a proven channel, and that the platforms are moving faster than most companies are organized to notice.
One update, one week, a different ChatGPT — leaving marketing teams scrambling
On August 8, ChatGPT switched its default model to GPT-5.6, and within a week it had changed how it researches an answer.
When you ask ChatGPT a question that needs current information, it searches the web behind the scenes before it replies. Until August, it almost always ran just one search: in July, 98% of conversations triggered a single search. By September, only 39% did. ChatGPT now usually breaks a question into several narrower searches, which changes which websites it finds and quotes. Its answers also draw on more of them: the average answer used 11 sources in July and 19 in August.
The searches themselves changed too. In July, almost none of ChatGPT's searches were limited to a specific website. By September, 21% were, as ChatGPT went straight to named sites instead of searching the open web. Searches for "best" products fell from 19% to 8%, while searches including "official" or the current year rose. In practice, ChatGPT stopped asking "what's the best X?" and started checking the brand's own site.
That changed which websites got cited. Peec AI compared each type of source with how often it was cited in an average week in Q2 2026. By mid-September, ChatGPT was citing 18% more sources overall than in that average week. Reddit went the other way: for every 100 times ChatGPT cited Reddit in a typical Q2 week, it cited it 59 times in late July and just 9 times by mid-September.
Brands' own pages filled the gap. By the end of Q3, ChatGPT was citing product pages two and a half times as often as in a typical Q2 week, and they rose from 13.8% to 20.1% of all the sources it cited. Company websites continued to make up more than half of everything ChatGPT cited (52% to 58% each week). "Best X" listicles were cited less often, and as other sources grew, their share halved from 16.2% to 7.3%.
There was no announcement and no migration period, and most marketing teams aren’t set up to notice a change like that when it happens. Almost half (47%) of marketers trace a lead, deal, or revenue to AI search at least weekly, but a traced deal isn’t proof. Only 24% can show what the channel returns overall. Where several teams share the work, 35% say it stalls while they wait on another team’s capacity. Marketers know the pace is a problem: “AI platforms are changing too quickly” is the barrier they name most often (33%), ahead of brand safety (30%) and a lack of reliable measurement (28%).
Marius Meiners, CEO at Peec AI, comments, “Marketing teams plan by the quarter and work toward annual targets. AI platforms ship changes on an engineering cadence, and one update can redraw the map in a week. The old playbook won’t keep up with that. AI search is quickly becoming one of the most important growth channels, and the teams pulling ahead measure continuously, adapt quickly and move together.”
Spending ahead of the proof
Marketing teams are investing anyway. For 61%, increasing visibility and recommendations by AI platforms is a top priority, and 60% have checked what an AI assistant says about their own brand several times in the past month. Over the past year, 89% raised AI search spend, and 81% expect to raise it again.
The pressure is coming from the market more than the boardroom. Marketers most often cite competitors appearing in AI answers (43%) and customers using AI tools to research or buy (42%), ahead of senior leadership or the board (34%) and industry media and hype (28%).
But proof hasn't kept pace with spend. Only 24% can already prove a return on AI search. Even among companies that raised spend significantly, 64% can't, and 54% of those who say they have reliable metrics they trust still can't show ROI.
What companies are doing, and what they're dropping
Even though AI search is new, a playbook is emerging that is already well-adopted. A majority are using the seven tactics in the survey: optimizing existing content (79%), shaping how AI describes the brand (77%), optimizing to be recommended by AI (75%), creating content written for AI (74%), making their sites easier for AI systems to read (69%), building a presence on Reddit, YouTube, and LinkedIn (67%), and earning mentions in third-party media and reviews (57%). But for every tactic, up to half of the teams using it are very confident it works (43% to 50%).
Most of the work is on content companies already have. In the past three months, 46% refreshed older content because of AI search, 43% produced more of certain formats, 39% changed the format or structure of existing content, and 36% changed what they write about. A quarter (25%) cut back some formats and 21% stopped producing some altogether. Only 14% changed nothing. That focus fits what ChatGPT now rewards: clear, current pages on a brand's own site.
When it comes to the tactics companies abandon, earning third-party mentions tops the list at 21%. That work usually belongs to PR, yet comms or PR teams own AI search at only 3% of companies. Building a presence on Reddit, YouTube, and LinkedIn comes next, dropped by 17%, in the same quarter ChatGPT sharply cut back on citing Reddit.
Whether a team keeps going depends on whether it can see results. Teams with reliable metrics they trust are far less likely to have dropped a tactic than teams with some visibility they can't quantify: 7% against 31% have stopped optimizing to be recommended by AI, and 8% against 25% have stopped creating content written for AI.
Organizational confusion holding AI search back
Part of the problem is organizational. AI search touches SEO, content, brand, PR, product, engineering, and analytics, and no single function owns it at most companies: the SEO team leads at 24% and executive leadership at 22%. Where more than one team is involved, 88% report coordination problems, most often conflicting priorities and metrics (39%) and waiting on other teams' capacity (35%).
Seniority changes the picture. Among owners and C-suite executives, 49% say they have metrics they trust, against 29% of the managers closest to the work.
The companies that can show results aren't the biggest. Mid-market and enterprise companies prove ROI at the same rate (24%). What sets them apart is how they have organized for AI search. Companies with an AI search strategy integrated into their wider marketing are the most likely to prove ROI (43%), and those with dedicated full-time AI search roles are twice as likely to have metrics they trust as those where it is part of a broader role (57% against 28%).
Integrated companies also stick with their tactics. Only 5% have stopped optimizing to be recommended by AI, against 31% of companies still exploring AI search. They bring more of the business into the work, too: 71% involve data and analytics teams, 44% involve product (against 24%) and 37% involve engineering (against 20%).
The US is further ahead
Marketers in the US, UK, and Germany are raising AI search budgets at the same rate (87%, 91%, and 89%), but American companies are getting more out of it. In the US, 31% of marketers can already prove a return, against 19% in the UK and 18% in Germany. US companies are also the most likely to have built AI search into their wider marketing strategy (38%, against 30% in the UK and 22% in Germany), and their customers are pushing harder: 47% cite customers using AI tools to research or buy as a reason to invest, against 38% in Germany. European teams give up on PR-led tactics faster. In Germany, 30% of marketers have stopped trying to earn third-party mentions, against 24% in the UK and 14% in the US.
The full report is available here.
About Primary Sources
Primary Sources is a quarterly research series from Peec AI. Each edition combines a survey of marketing decision-makers with Peec AI's platform data to report where what marketers believe and plan diverges from what AI platforms actually do. The Q3 2026 survey was conducted by Sapio Research among 700 marketing decision-makers, manager level and above, at companies with 100+ employees in the US (300), UK (200), and Germany (200), in September 2026. Peec AI analyzed AI search behavior across ChatGPT, Google AI Overviews, Google AI Mode, Gemini, and Claude from July–September 2026, with selected weekly trends extending into early October. Citation-volume indexes use the Q2 2026 average as a baseline of 100.
Press Inquiries
Noah Wolff
growth [at] peec.ai
https://peec.ai
