Named,
Not Cited

Questel is a large, well-covered brand, so the honest reading is not “you are invisible.” You appear in most answers. The exploitable gap is one level down: on your own highest-intent queries, the sources AI cites are owned by competitors, and questel.com supplies almost none of them. That is a source-ownership problem, which an off-site content and citation program can move, not a brand-awareness problem you already know about.

Every item below is a verified gap against competitors, paired with the concrete work that would close it. These are levers, not guarantees: the work is measurable and ownable, and today it is unowned.

Across the whole scan, the single most-cited vendor domain is a direct competitor, cypris.ai (101), ahead of questel.com (80). Gemini’s citations are opaque Google proxies and are excluded; ownership rests on Perplexity, Claude, and ChatGPT real URLs.

Zoomed into your core: patent search & FTO

This is the cluster Questel is betting its new AI on, and the one where the citation gap is total: questel.com is cited 0 of 435 times, while competitor domains supply ~30% of sources. questel.com is cited 80 times elsewhere in the scan, so this zero is a real category gap, not a scan artifact.

Confirmed on a fifth surface: Google’s AI

We re-ran the same buyer keywords through Google AI Overviews and AI Mode, a surface with no prior data. Identical pattern: when Google’s AI answer fired, Questel was named 15 times but questel.com was cited once, and the most-cited competitor domain was again Cypris. On “freedom to operate analysis” specifically, Google’s AI answer fires, cites Cypris, and does not name Questel at all.

Five surfaces now agree: ChatGPT, Claude, Perplexity, Gemini, and Google AI all name Questel and cite competitors, led by Cypris, over questel.com. This is not one engine’s quirk.

Aggregate counts are the evidence; these are what they look like on the page. Real excerpts from the grounded runs, unedited.

Where you win ChatGPT · “best patent search software for large companies” “2. Questel Orbit Intelligence — Best for: Freedom-to-operate (FTO), landscape and competitive intelligence with strong dashboards. Powerful search (keyword, classification, semantic), good family/grouping logic and status indicators.”

Questel is ranked and recommended. Named 11/12 on this query. The engine can place you well when the sourcing supports it.

The gap Perplexity · “Questel vs PatSnap” “PatSnap and Questel are both strong patent/IP intelligence platforms... Questel is broader and more oriented to the full IP lifecycle, including patents, trademarks, designs, domains, and portfolio management.”

You are described fairly and in full. But every source under the answer is a third-party comparison site or a competitor. questel.com is not one of them. Verified live below.

The gap Claude · “AI patent search tools” “AI patent search tools have transformed how inventors, IP professionals, and R&D teams find prior art... these platforms use semantic search, embeddings, and large language models.”

Questel is absent from this answer. The tools named and cited are Cypris, DeepIP, PatSnap, IPRally, NLPatent, PatSeer, the AI-native challengers, on the exact query your QaECTER launch targets.

Stale Perplexity · “what products does Questel offer” “Innovation management software, including innosabi products for idea management and innovation intelligence.”

innosabi was divested to Collaboration.Ai in January 2026. AI still lists it as a current Questel product in 7/12 runs, with no divestiture language.

Sentiment Claude · “Questel reviews” “...their firm subscribed to Questel/Markify’s trademark watch service in August 2025, but despite repeated follow-ups, the company failed to deliver the first monthly report.”

This single Trustpilot complaint is the recurring lead “con” in 9 of 9 Claude runs. When AI judges you, ~67% of its sources are reputation sites like this, not software-review sites.

This is the guard against overclaiming. Questel is present far too often to call it “invisible,” and that is important: it keeps every gap claim honest and targetable. The gaps are specific and often platform-specific, which is exactly what makes them addressable.

Read the rows: “IP management software,” the highest-demand term, is absent on ChatGPT and Claude (0/3 each). “AI patent search tools” is absent on Claude and Gemini. “patent analytics software” is a clean 12/12. The pattern is uneven coverage, not absence, which is a content-and-sourcing problem, not a lost cause.

Two things frame the opportunity. First, Questel is narrowing to its IP-legal core, not expanding: it built an innovation stack then sold it off, and every new AI bet is IP-legal. Second, the R&D and patent-search ground it is retreating toward and betting on is exactly where AI-native rivals have planted their content flag.

• Divested innosabi (innovation mgmt) to Collaboration.Ai, Jan 2026, and doeLEGAL in 2024, both under “sharpen focus on IP” framing.
• Every new AI investment is IP-legal: qatent (drafting), Sophia (Oct 2025), and QaECTER, a semantic patent-search model launched April 2026.
• CEO Frederic Beylier (since Mar 2026): the goal is “market leader for AI solutions for IP,” buyer named as corporate IP departments and law firms.
• The irony that makes opportunity 02 sharp: Questel shipped an AI patent-search model in April, yet on “how do I search patents” its AI line is cited 0 of 36 times while IPRally and Patlytics are named.

Who owns each buyer’s front door

AI leans on software-review sites for “best” and “alternatives” questions. Questel is thin on the exact surface AI cites most, and that surface is consolidating under one owner, which makes a focused review program unusually leveraged right now.

Honest scope: the thinness is G2-specific. On Capterra (Orbit 13) and Gartner PI (2, the leader) Questel is fine. Do not present this as “no reviews anywhere.”

Where real buyers search (worldwide monthly, DataForSEO). This is why the scan centered on patent-search and IP-management, and why the R&D cluster was cut: it is near-empty everywhere. Branded demand is the backdrop, not a hook.

Branded search, worldwide monthly

PatSnap pulls 2.2x Questel’s worldwide branded search and 4x in the UK, but Questel owns France ~5x. Context for the room, not a finding to lead with.

Every fact re-derived from raw JSON and put through an adversarial gate: independent re-computation, a live-recheck survivability test, and a backfire check. Mechanism facts lead.

• Grounded only. 4 engines (ChatGPT, Claude, Perplexity, Gemini) in live-search mode, 3 runs each, 336 answers, citations captured. Frequencies, never absolutes.
• Re-derived, not summarized. Every number recomputed from raw JSON. 42 candidate facts, 31 survived a three-part adversarial gate.
• Backfire-tested. Anything that would collapse when an in-house lead types the query herself was cut. The presence map is the guard: Questel appears too often to ever be called invisible.