AI SEO Statistics: Patent Broker (2026-07 edition)
40 questions · 120 AI responses · 3 models · measured 2026-07-06
The question bank
The questions we tested — sampled from real buyer journeys in patent broker.
Each model answered every question once, same wording, same day. These are the prompts behind every percentage on this page.
Show all 40 questions
Model by model
23-point average divergence: which AI you ask changes the answer.
The divergence index is the average gap between the most and least likely model per behavior. Higher = the models disagree more about patent broker buyers.
| ChatGPT | Claude | Gemini | Consensus | |
|---|---|---|---|---|
| Recommends hiring a professional | 88% | 68% | 43% | 45% |
| Suggests DIY first | 15% | 13% | 13% | 73% |
| Names specific providers | 20% | 25% | 20% | 68% |
| Gives price or cost info | 18% | 23% | 35% | 60% |
| Tells to check reviews | 8% | 10% | 0% | 85% |
| Tells to verify credentials | 10% | 18% | 0% | 75% |
| Mentions case studies / portfolio | 15% | 28% | 3% | 68% |
| Mentions local proximity | 3% | 3% | 0% | 95% |
| Gives selection criteria | 28% | 45% | 23% | 50% |
| Warns about red flags | 15% | 20% | 8% | 78% |
| Asks a clarifying question | 50% | 78% | 0% | 10% |
| Recommends multiple quotes | 10% | 8% | 0% | 85% |
By model
How each assistant handled Patent Broker questions.
Reading the 120 answers model by model shows how differently the three assistants treat the same patent broker questions. On the most consequential behavior — whether to send the buyer to a professional at all — the rate ranged from 87.5% (ChatGPT) down to 42.5% (Gemini), a 45-point gap on an identical question set.
Across the 40 patent broker answers it produced, ChatGPT recommended hiring a professional in 87.5% of them and suggested a DIY approach first 15% of the time. It named a specific provider in 20% of answers (about 0.6 distinct providers per answer) and included price or cost information 17.5% of the time. ChatGPT asked a clarifying question before answering in 50% of cases, warned about red flags or scams in 15%, and told the buyer to verify credentials in 10%, averaging 575 words per answer. On the remaining cues it told the buyer to check reviews in 7.5%, pointed to case studies or a portfolio in 15%, and framed the choice around local proximity in 2.5%; a selection-criteria checklist appeared in 27.5% of its answers and a recommendation to gather multiple quotes in 10%.
Across the 40 patent broker answers it produced, Claude recommended hiring a professional in 67.5% of them and suggested a DIY approach first 12.5% of the time. It named a specific provider in 25% of answers (about 1.2 distinct providers per answer) and included price or cost information 22.5% of the time. Claude asked a clarifying question before answering in 77.5% of cases, warned about red flags or scams in 20%, and told the buyer to verify credentials in 17.5%, averaging 312 words per answer. On the remaining cues it told the buyer to check reviews in 10%, pointed to case studies or a portfolio in 27.5%, and framed the choice around local proximity in 2.5%; a selection-criteria checklist appeared in 45% of its answers and a recommendation to gather multiple quotes in 7.5%.
Across the 40 patent broker answers it produced, Gemini recommended hiring a professional in 42.5% of them and suggested a DIY approach first 12.5% of the time. It named a specific provider in 20% of answers (about 0.6 distinct providers per answer) and included price or cost information 35% of the time. Gemini asked a clarifying question before answering in 0% of cases, warned about red flags or scams in 7.5%, and told the buyer to verify credentials in 0%, averaging 278 words per answer. On the remaining cues it told the buyer to check reviews in 0%, pointed to case studies or a portfolio in 2.5%, and framed the choice around local proximity in 0%; a selection-criteria checklist appeared in 22.5% of its answers and a recommendation to gather multiple quotes in 0%.
Taken together, ChatGPT is the assistant most likely to route a patent broker buyer to a professional (87.5%) and Gemini the least (42.5%). ChatGPT produced the longest answers, at 575 words on average. Specific providers were named most often by Claude (25%) — even there, roughly one answer in 4 carried a name.
Where they disagree
The behaviors where the choice of model changes the answer.
The divergence index for this study is 22.8 points — the average distance between the most and least likely model across the coded behaviors. The gaps below are where which assistant a patent broker buyer happens to ask matters most:
- Asks a clarifying question: from 0% (Gemini) to 77.5% (Claude) — a 78-point spread.
- Recommends hiring a professional: from 42.5% (Gemini) to 87.5% (ChatGPT) — a 45-point spread.
- Mentions case studies or portfolio: from 2.5% (Gemini) to 27.5% (Claude) — a 25-point spread.
- Gives selection criteria: from 22.5% (Gemini) to 45% (Claude) — a 23-point spread.
- Gives price or cost information: from 17.5% (ChatGPT) to 35% (Gemini) — a 18-point spread.
The widest single gap — asks a clarifying question, 78 points — means a patent broker buyer can receive materially different guidance on the same question depending only on which assistant they happen to open, so any visibility strategy built on a single model's behavior describes only part of the patent broker market.
Where they agree
The points of near-consensus in Patent Broker.
On other behaviors the three models move almost in lockstep — the points of near-consensus for patent broker, where all three landed within a few points of each other:
- Suggests a DIY approach first: 12.5%–15% across all three (a 3-point spread).
- Mentions local proximity: 0%–2.5% across all three (a 3-point spread).
- Names a specific provider: 20%–25% across all three (a 5-point spread).
- Tells the buyer to check reviews: 0%–10% across all three (a 10-point spread).
Measured question by question, the three assistants coded a response the same way most consistently on "mentions local proximity" (identical coding in 95% of questions) and least consistently on "asks a clarifying question" (10%).
Every behavior, measured
All twelve coded behaviors for Patent Broker, averaged across the three models.
The behaviors AI models reproduce most often for patent broker are recommends hiring a professional (65.8% on average), asks a clarifying question (42.5%) and gives selection criteria (31.7%); the rarest are mentions local proximity (1.7%), recommends multiple quotes (5.8%) and tells the buyer to check reviews (5.8%). Each figure below is the share of a model's 40 answers in which the behavior appeared at least once, averaged across the 3 models with the full per-model range in parentheses:
- Recommends hiring a professional: 65.8% on average (ChatGPT 87.5%, Claude 67.5%, Gemini 42.5%) — a 45-point spread.
- Asks a clarifying question: 42.5% on average (ChatGPT 50%, Claude 77.5%, Gemini 0%) — a 78-point spread.
- Gives selection criteria: 31.7% on average (ChatGPT 27.5%, Claude 45%, Gemini 22.5%) — a 23-point spread.
- Gives price or cost information: 25% on average (ChatGPT 17.5%, Claude 22.5%, Gemini 35%) — a 18-point spread.
- Names a specific provider: 21.7% on average (ChatGPT 20%, Claude 25%, Gemini 20%) — a 5-point spread.
- Mentions case studies or portfolio: 15% on average (ChatGPT 15%, Claude 27.5%, Gemini 2.5%) — a 25-point spread.
- Warns about red flags or scams: 14.2% on average (ChatGPT 15%, Claude 20%, Gemini 7.5%) — a 13-point spread.
- Suggests a DIY approach first: 13.3% on average (ChatGPT 15%, Claude 12.5%, Gemini 12.5%) — a 3-point spread.
- Tells the buyer to verify credentials: 9.2% on average (ChatGPT 10%, Claude 17.5%, Gemini 0%) — a 18-point spread.
- Tells the buyer to check reviews: 5.8% on average (ChatGPT 7.5%, Claude 10%, Gemini 0%) — a 10-point spread.
- Recommends multiple quotes: 5.8% on average (ChatGPT 10%, Claude 7.5%, Gemini 0%) — a 10-point spread.
- Mentions local proximity: 1.7% on average (ChatGPT 2.5%, Claude 2.5%, Gemini 0%) — a 3-point spread.
Trust signals
How well the models protect the patent broker buyer.
Beyond whether to hire, the rubric codes how carefully each assistant protects the patent broker buyer once a decision is made. Telling the buyer to check reviews or ratings appeared in 5.8% of answers on average. Verifying credentials or certifications appeared in 9.2%. Warning about red flags or scams appeared in 14.2%.
On structuring the decision, a selection-criteria checklist showed up in 31.7% of answers on average and a recommendation to gather multiple quotes in 5.8%. The single least-reproduced protective signal for patent broker is "tells the buyer to check reviews" at 5.8% on average — the clearest opening for content that supplies it, since the models are not yet reliably surfacing that guidance on their own.
Referral behavior
Do AI models name Patent Broker providers?
For service providers the decisive question is whether these systems name anyone at all. Across 120 patent broker answers, a specific provider was named in 21.7% of responses on average — roughly 0.8 distinct providers per answer. In practice the assistants behave far more as an explanatory layer than as a referral engine for patent broker: visibility comes from being the reasoning a model reproduces, not from being the named recommendation.
The question set
What these 40 Patent Broker questions cover.
The 40 questions behind every percentage on this page were drawn from real patent broker (legal services; buyer hiring decisions for this specific service) buyer journeys. Each was put to all 3 models once, with identical wording, so the rates above describe how the assistants handled this exact patent broker question set — not a general prior or a hand-picked subset. The full list is shown earlier on this page; the coded percentages are what those specific questions produced.
How to read this
A note on the numbers.
A percentage here is the share of a model's 40 answers in which the behavior appeared at least once — not a confidence score. Because each model answered every question exactly once on 2026-07-06, the figures describe this specific patent broker question set and snapshot rather than a general prior. The full protocol and coding rubric are documented in the study methodology.
Methodology
A controlled snapshot, documented end to end.
40 standardized buyer questions per industry, one response per model per question (ChatGPT (gpt-5-mini), Claude (claude-sonnet-5), Gemini (gemini-3-flash-preview)), collected 2026-07-06, coded against a fixed 12-behavior rubric with human QA. AI outputs vary with model version, location and time — figures describe this sample and window, and are refreshed each edition. Read the full methodology →