The AI Discovery Crisis: How Competitors and Publishers Are Hijacking Your B2B Product Narrative

By Global Tech & Marketing Desk
Published: September 2026

Main Facts: The New Reality of AI-Driven B2B Discovery

In the modern enterprise software market, the traditional buyer’s journey—spanning search engine result pages (SERPs), software review directories, and vendor websites—has fundamentally shifted. Today, prospective enterprise buyers increasingly rely on generative artificial intelligence engines like ChatGPT, Google AI Overviews, Claude, Gemini, and Perplexity to evaluate, contrast, and select business-to-business (B2B) software.

However, a groundbreaking empirical study reveals an uncomfortable truth for software vendors: AI-powered discovery platforms rarely rely on first-party vendor content when comparing competing products.

According to recent data from a comprehensive study conducted by Foundation and AirOps, which analyzed 240 AI-generated answers across six major AI platforms using 20 distinct Applicant Tracking System (ATS) buying prompts, direct-party vendor content is largely sidelined. Across six direct matchups among leading HR tech platforms—Greenhouse, Ashby, Lever, and Workable—first-party comparison pages accounted for a meager 27% of visible source presence.

The remaining 73% of the evidence used by AI engines to define product features, capabilities, and shortcomings came from external sources. These include independent commercial publishers, third-party software review aggregators, community forums, and—most critically—direct competitors.

If a software vendor fails to proactively construct and optimize its own comparison and category ecosystem, third parties—and sometimes direct rivals—will control the narrative presented to high-intent buyers at the point of decision.

Chronology: How the "Versus Layer" Hijacked Search and AI

The erosion of first-party brand control in organic search has been underway for a decade, accelerated by the rise of programmatic SEO, aggressive review aggregators, and publisher listicles. However, the transition to Generative Engine Optimization (GEO) has compressed this vulnerability into a critical operational threat.

• Phase 1: The Traditional SERP Era (Pre-2024): B2B buyers relied on Google keyword searches such as "[Vendor A] vs [Vendor B]." Vendors could easily rank for these terms with simple landing pages, neutralizing competitor positioning.
• Phase 2: The Aggregator Boom (2024–2025): As search engines integrated rich snippets and AI summaries, third-party publishers and review sites (like G2, Capterra, and specialized industry blogs) dominated the top of search results, capturing high-intent organic traffic.
• Phase 3: The Generative AI Era (2026): Platforms like ChatGPT, Google AI Overviews, and Claude stopped serving raw lists of links entirely. Instead, they synthesized answers by crawling the web for structured comparisons. Crucially, because AI looks for objective, multi-perspective evidence to build balanced responses, it began pulling heavily from third-party ecosystems—including competitor-authored comparison pages.

This evolutionary timeline culminated in the 2026 findings, where automated AI research tools demonstrated a systemic reliance on what industry analysts have dubbed the "Versus Layer."

Supporting Data: Dissecting the 73% External Source Phenomenon

To understand how AI platforms construct product comparisons, the Foundation research team tracked 2,695 recorded source presences across 240 AI responses generated by ChatGPT, Perplexity, Gemini, Claude, Google AI Mode, and Google AI Overview.

The findings upend conventional wisdom regarding domain authority, backlink profiles, and SEO strategies.

1. The Breakdown of the "Versus Layer"

When AI engines look outside the primary vendors to explain product differences, the distribution of sources breaks down as follows:

• Independent & Commercial Publishers: 42.7%
• Other Software Vendors (Competitors): 20.0%
• Software Review Sites: 7.4%
• Community Sources (e.g., Reddit, StackExchange): 2.9%
• First-Party Vendor Content: 27.0%

2. Domain Authority Is No Longer a Deciding Factor

Traditional SEO dogma dictates that high Domain Rating (DR) or URL Rating (UR) paired with extensive backlink profiles determines what ranks. AI discovery operates differently.

For instance, when evaluating a matchup between Greenhouse and Ashby, AI utilized a comparison page published by Lever (a direct competitor) in 27 distinct answers. By contrast, Ashby’s own first-party page on the same topic was cited only 18 times.

According to Ahrefs data, Lever’s comparison URL possessed a URL Rating of 0, a meager 2 referring domains, and a paltry 8 monthly visits. Meanwhile, Ashby’s page had a UR of 19, 6 referring domains, and roughly 20 monthly visits. The data proves that AI engines prioritize topical relevance and structural clarity over raw backlink authority when synthesizing answers.

3. The Buyer Journey Dictates AI Source Preferences

The type of source an AI engine pulls from shifts dramatically depending on where the user is in the buying journey:

• Neutral Category Prompts (Top-of-Funnel): Listicles and broad roundups accounted for 51.1% of surface-weighted source presence, while dedicated comparison pages accounted for just 12.6%.
• Vendor-Involved Prompts (Mid-Funnel): Once a specific vendor name entered the query, comparison pages surged to 33.8%, while listicles dropped to 21.4%.
• Direct Head-to-Head Questions (Bottom-of-Funnel): When asked to directly contrast two specific products, dedicated comparison pages dominated, reaching 42.1% of source presence.

4. Structure Over Marketing Fluff

The structural layout of a comparison page dictates how frequently an AI model will ingest and cite it. Pages that surfaced buyer fit, explicit decision criteria, trade-offs, and clear recommendations early in the text were cited exponentially more often.

For example, a third-party page by Venture Harbour comparing Workable and Greenhouse immediately breaks down ideal customer profiles, evaluation criteria, and core trade-offs. It appeared in 13 AI answers. Conversely, Workable’s own official comparison page for the same matchup—which spent its introductory paragraphs highlighting product benefits and high-level demo messaging—appeared in just 2 answers.

5. The ChatGPT Anomaly

While platforms like Gemini, Claude, Perplexity, and Google AI Overviews lean heavily on the Versus Layer, ChatGPT behaves fundamentally differently.

As the most widely used consumer AI platform in the United States, ChatGPT relies disproportionately on first-party product and support documentation. In tests comparing mid-market recruiting tools Lever and Workable, ChatGPT’s citations pointed primarily to Lever’s product/pricing pages and Workable’s help documentation.

Product and support content accounted for 77.3% of visible source presence on ChatGPT (with 63.8% originating strictly from product pages), while dedicated comparisons and listicles accounted for a negligible 3.3% combined.

Official Responses and Platform Shifts

As empirical evidence mounts regarding the shifting dynamics of AI search traffic, software vendors and intelligence firms are responding.

Comscore’s Q2 2026 AI Intelligence Report highlights a major decentralization in user prompt volume across platforms. While ChatGPT remains the market leader, its market share of AI prompts dropped from 70% to 50% over the past year. Simultaneously, Google’s Gemini surged from 17% to 30%, and Anthropic’s Claude jumped from 2% to 11%.

Because these distinct engines crawl and prioritize different datasets—with Google ecosystems relying on broader web indexes, ChatGPT indexing deep product/support architectures, and Claude favoring structured analytical text—B2B marketing leaders can no longer rely on a monolithic SEO strategy.

Industry analysts emphasize that software CMOs must audit their digital footprints across multiple generative architectures to ensure brand narratives are not being co-opted by competitors writing biased or inaccurate comparison pages.

Implications for B2B Marketers and Product Leaders

The realization that 73% of AI-driven product comparisons are controlled by outside parties carries profound strategic implications for software companies.

1. The Risk of Competitor-Authored Narratives

When a company lacks a robust, well-structured first-party comparison page, AI engines have no choice but to source their information elsewhere. As demonstrated in the Greenhouse-Ashby case study, if you do not define your product’s limitations and strengths, your competitors will gladly do it for you—potentially highlighting weaknesses, omitting crucial features, or framing pricing models in an unfavorable light.

2. A Paradigm Shift in Content Strategy

Traditional content marketing has long prioritized top-of-funnel broad traffic (e.g., "What is an ATS?"). However, in an AI-first search environment, bottom-of-funnel and mid-funnel battleground pages—specifically competitor comparisons and objective trade-off analyses—are the primary fuel for generative answers. Marketers must pivot resources toward building exhaustive, unbiased, and structurally optimized comparison hubs.

3. Tactical Recommendations for B2B Brands

To reclaim control of AI discovery channels, marketing and product teams should execute a three-step auditing and optimization framework:

• Audit the "Versus Layer": Identify every possible head-to-head combination of your product versus your top five competitors. Query major AI platforms (ChatGPT, Claude, Gemini, Perplexity) to see what sources currently dominate the answers.
• Restructure Comparison Architecture: Rewrite owned comparison pages to place buyer fit, evaluation criteria, and trade-offs in the immediate introduction. Avoid overly promotional marketing fluff or gated demo forms that prevent AI web scrapers from reading core text.
• Diversify Platform Optimization: Recognize that ChatGPT favors deep product and documentation pages, whereas Google and Perplexity rely heavily on third-party publishers and comparison tables. Ensure technical documentation is clean, public, and easily indexable.

Conclusion

The era of renting your brand narrative to review sites and third-party publishers is over. In the age of AI discovery, the rule is absolute: If you do not supply the evidence about your own product, somebody else will—and they may not be on your side.