Your Competitors Are Not Who Google Says They Are

The Short Answer

The companies ranking on page one of Google are not the same companies AI recommends to your buyers. Only 14% of URLs overlap between Google AI Mode results and Google's organic top 10. That means 86% of the sources AI cites when answering buyer questions are companies that do not rank in traditional search. Your competitive set has shifted. Most companies are still watching the old one.

The Competitive Set You Are Tracking No Longer Exists

Every enterprise marketing team has a competitive watchlist. It is usually built from three inputs: SEO keyword rankings, analyst reports, and deal losses in CRM. All three pull from the same pool. The companies you already know about.

That made sense when Google was the primary research channel. Rank for the right keywords, monitor who else ranks, adjust positioning. Clean loop.

The loop broke.

SE Ranking's 2025 study analysed thousands of queries across Google AI Mode and found that AI Mode's domain overlap with organic top 10 results is only 21.9%. Not a minor deviation. A structural divergence. The sources AI pulls from to answer category-level questions -- "best enterprise CRM for mid-market," "top UK accountancy software" -- are different sources than the ones ranking in traditional search.

This is not an edge case. It is the new baseline.

Why AI Recommends Different Companies

Google's organic algorithm rewards signals that have compounded for years: backlink authority, domain age, technical SEO, content volume. These favour incumbents. Companies that have been publishing the longest tend to rank the highest.

AI recommendation works on a different architecture. It synthesises across sources -- review platforms, comparison content, community discussions, analyst commentary, third-party guides. It looks for where a company name appears in context that matches the query intent, not where a company's own website ranks.

This creates a different competitive surface. A company that appears in 12 buyer guides, 6 comparison threads, and 3 analyst evaluations will be cited by AI even if its website sits on page four of Google. A company with a page-one ranking but no third-party recommendation content may not appear in AI responses at all.

Our research confirms this pattern. Brands are 6.5x more likely to be cited via third-party sources than through their owned domain. The citation signal is not coming from your website. It is coming from everywhere else.

The Data That Makes This Concrete

Three datasets converge on the same finding.

The URL gap. SE Ranking found 86% of AI-cited sources are NOT in Google's top 10 organic results. The overlap is so thin that optimising for one channel gives you almost no visibility in the other.

The buyer behaviour shift. Forrester's 2025 B2B buyer study found that 57% of B2B buyers now consider more or different vendors due to generative AI. Not the same vendors surfaced through different channels. Different vendors entirely. AI is introducing companies that were never in the buyer's original consideration set.

The shortlist effect. Vendors mentioned by all major AI platforms are 3.2x more likely to be shortlisted by buyers. And according to 6sense, 95% of winning vendors were on the buyer's Day 1 shortlist. If AI is building that shortlist and your competitors are not who you think they are, you are optimising against the wrong field.

What This Means for Your Strategy

The implication is not that Google rankings stopped mattering. They still drive traffic. The implication is that your competitive intelligence is incomplete.

You are likely tracking 10-15 companies. The companies appearing in AI recommendations for your category keywords may include 8-10 you have never monitored. Some may be smaller. Some may be in adjacent categories. Some may be companies you dismissed years ago.

This is the delta most teams are missing.

Our AI Visibility Benchmark at 150 companies found that 81% of B2B companies have no meaningful AI citation presence. They are invisible to the channel that is now shaping buyer shortlists. But the 19% who are visible -- they are the new competitive set. And they may not overlap with the names on your existing watchlist.

What to Do About It

1. Run a competitive audit through AI, not just Google. Ask ChatGPT, Perplexity, Gemini, and Google AI Mode your category-level buying queries. "Best [your category] for [your ICP]." Record which companies appear. Compare that list against your current competitive watchlist. The delta is your blind spot.

2. Map your third-party citation footprint. Your owned website is one signal. What matters more is where your company name appears across review sites, comparison content, buyer guides, community threads, and analyst commentary. If you are absent from those surfaces, AI has nothing to cite. Our AI Visibility Audit maps this across all four dimensions.

3. Redefine what "competitive" means. A competitor is no longer just a company that bids on the same keywords or appears in the same analyst quadrant. A competitor is any company that AI recommends when a buyer asks a question you should be answering. That list is wider, stranger, and more dynamic than what most teams are tracking.

4. Build citation infrastructure, not just content. The fix is not publishing more blog posts on your website. The fix is appearing in the places AI pulls from -- third-party sources, comparison content, community discussions. This is AI visibility infrastructure, not traditional SEO.

5. Monitor the shift quarterly. The AI competitive landscape moves faster than organic rankings. A company that was invisible three months ago can appear in AI recommendations after a single well-placed analyst report or buyer guide. Build a quarterly review cadence. Our Competitive Intelligence Report tracks your position against AI-era competitors across all major platforms.

---