Vertical Intelligence Reports
One report per vertical: who the AI models recommend, how the leaders differ by region, and what the local winners are doing right. Eleven verticals from the July 2026 edition.
What's inside
- The vertical podium across all eight AI models
- Regional splits: where the national leader loses locally
- The visibility patterns the winners share
Get the paper
- Branding Agencies (PDF)
- Content Marketing Agencies (PDF)
- Digital Marketing Agencies (PDF)
- HVAC Companies (PDF)
- Law Firms (PDF)
- PR Agencies (PDF)
- Roofing Companies (PDF)
- SEO Companies (PDF)
- Social Media Agencies (PDF)
- Staffing Agencies (PDF)
- Web Design Companies (PDF)
Eleven reports from the July 2026 edition. Read any of them on this page or download the PDF.
Read the reports
Branding Agencies
When buyers ask the 8 leading AI models for branding agencies, these are the names they hear. Measured from the question a real buyer asks, captured verbatim, frozen as the July 2026 record.
- Pentagram holds #1 this edition, named by 6 of 8 AI models. ChatGPT and Gemini do not name it at all.
- 3/10 of the top ten are named by a majority of AI models. The rest hold thinner consensus.
- 2/4 regional editions crown a different #1 than the global ranking. Ask locally, get a different answer.
The top ten, per AI model
Filled dot = that AI model recommends the brand unprompted when asked "What are the best branding agencies?"
| # | Brand | Score | Named by | ChatGPT | Claude | Gemini | Perplexity | DeepSeek | Grok | Copilot | Google AI |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Pentagram (pentagram.com) | 47 | 6/8 | No | Yes | No | Yes | Yes | Yes | Yes | Yes |
| 2 | Wolff Olins (wolffolins.com) | 33.2 | 5/8 | No | No | Yes | Yes | Yes | Yes | Yes | No |
| 3 | Landor (landor.com) | 33 | 5/8 | No | Yes | No | Yes | Yes | Yes | No | Yes |
| 4 | Clay (clay.com) | 26.6 | 4/8 | Yes | No | Yes | Yes | No | No | Yes | No |
| 5 | Windows (microsoft.com) | 26.5 | 4/8 | No | Yes | Yes | No | Yes | Yes | No | No |
| 6 | Siegel+Gale (siegelgale.com) | 26.2 | 4/8 | Yes | No | No | Yes | No | Yes | Yes | No |
| 7 | COLLINS (collins1.com) | 20.6 | 3/8 | No | No | Yes | No | Yes | No | Yes | No |
| 8 | Mastercard (mastercard.com) | 20.3 | 3/8 | No | Yes | No | No | Yes | Yes | No | No |
| 9 | Spotify (spotify.com) | 19.7 | 3/8 | No | No | Yes | No | No | Yes | No | Yes |
| 10 | Mission Control (missioncontrol.com) | 19.7 | 3/8 | No | No | No | Yes | No | Yes | Yes | No |
AI Recommendation Score 0 to 100; full published formula at citehawk.com/leaderboards/methodology. Snapshot d4c7db2a, captured 2026-07-02, immutable.
Cross a border, get a different shortlist
| Asked from | #1 in that market | vs global |
|---|---|---|
| United States | Pentagram | Matches global |
| United Kingdom | Pentagram | Matches global |
| Australia | Hulsbosch | Different #1 |
| Canada | Cossette | Different #1 |
Regional editions are published only where the local ranking meaningfully differs from global. Visibility earned in one market does not automatically transfer.
What to do with this record
If you are on the list
Your position is per-model, and the dot grid shows exactly where it leaks. Each empty dot is one AI model answering your buyers without you, and each has specific causes: the sources that AI model leans on do not carry you, or describe you wrongly. Claiming your brand on the Index is free, never affects position, and adds movement alerts the moment any AI model changes its answer.
If you are not on the list
The 8 AI models did not name you when buyers asked. That is an invisibility problem with a known playbook: verify what each AI model currently says about you, correct the wrong facts at their sources, and win presence on the sources these answers cite. Pull the receipts on the live category page and list who gets cited alongside the ranked brands; that list is the work.
The cadence that matters
This record re-freezes monthly, and the underlying answers are re-measured weekly on the platform. Rankings move as models retrain and retrieval refreshes, which is why a position is a feed to watch, not a trophy to file.
Content Marketing Agencies
When buyers ask the 8 leading AI models for content marketing agencies, these are the names they hear. Measured from the question a real buyer asks, captured verbatim, frozen as the July 2026 record.
- Fractl holds #1 this edition, named by 6 of 8 AI models. Claude and Grok do not name it at all.
- 4/10 of the top ten are named by a majority of AI models. 3 hold a ranking on the two-model minimum: fragile positions.
- 4/4 regional editions crown a different #1 than the global ranking. Ask locally, get a different answer.
The top ten, per AI model
Filled dot = that AI model recommends the brand unprompted when asked "What are the best content marketing agencies?"
| # | Brand | Score | Named by | ChatGPT | Claude | Gemini | Perplexity | DeepSeek | Grok | Copilot | Google AI |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Fractl (fractl.com) | 49 | 6/8 | Yes | No | Yes | Yes | Yes | No | Yes | Yes |
| 2 | Siege Media (siegemedia.com) | 46.4 | 5/8 | Yes | Yes | Yes | No | No | Yes | Yes | No |
| 3 | Brafton (brafton.com) | 40.7 | 5/8 | Yes | Yes | Yes | Yes | No | No | Yes | No |
| 4 | Animalz (animalz.co) | 39.8 | 5/8 | No | No | Yes | Yes | Yes | Yes | No | Yes |
| 5 | Contently (contently.com) | 32 | 4/8 | No | No | Yes | Yes | Yes | No | No | Yes |
| 6 | Shutterfly (shutterfly.com) | 24.7 | 3/8 | Yes | Yes | Yes | No | No | No | No | No |
| 7 | ClearVoice (clearvoice.com) | 24.2 | 3/8 | No | No | Yes | Yes | No | Yes | No | No |
| 8 | Grow and Convert (growandconvert.com) | 18.6 | 2/8 | No | No | No | Yes | No | Yes | No | No |
| 9 | Column Five (columnfivemedia.com) | 18 | 2/8 | Yes | No | No | Yes | No | No | No | No |
| 10 | Asana (asana.com) | 17.1 | 2/8 | Yes | No | Yes | No | No | No | No | No |
AI Recommendation Score 0 to 100; full published formula at citehawk.com/leaderboards/methodology. Snapshot cbbc63fd, captured 2026-07-02, immutable.
Cross a border, get a different shortlist
| Asked from | #1 in that market | vs global |
|---|---|---|
| United States | Siege Media | Different #1 |
| United Kingdom | Brafton | Different #1 |
| Australia | Brafton | Different #1 |
| Canada | Let's Get Optimized | Different #1 |
Regional editions are published only where the local ranking meaningfully differs from global. Visibility earned in one market does not automatically transfer.
What to do with this record
If you are on the list
Your position is per-model, and the dot grid shows exactly where it leaks. Each empty dot is one AI model answering your buyers without you, and each has specific causes: the sources that AI model leans on do not carry you, or describe you wrongly. Note that positions held on the two-model minimum are the most fragile on the board: one AI model changing its mind removes the ranking entirely. Claiming your brand on the Index is free, never affects position, and adds movement alerts the moment any AI model changes its answer.
If you are not on the list
The 8 AI models did not name you when buyers asked. That is an invisibility problem with a known playbook: verify what each AI model currently says about you, correct the wrong facts at their sources, and win presence on the sources these answers cite. Pull the receipts on the live category page and list who gets cited alongside the ranked brands; that list is the work.
The cadence that matters
This record re-freezes monthly, and the underlying answers are re-measured weekly on the platform. Rankings move as models retrain and retrieval refreshes, which is why a position is a feed to watch, not a trophy to file.
Digital Marketing Agencies
When buyers ask the 8 leading AI models for digital marketing agencies, these are the names they hear. Measured from the question a real buyer asks, captured verbatim, frozen as the July 2026 record.
- Ignite Visibility holds #1 this edition, named by 6 of 8 AI models. ChatGPT and DeepSeek do not name it at all.
- 2/10 of the top ten are named by a majority of AI models. 6 hold a ranking on the two-model minimum: fragile positions.
- 4/4 regional editions crown a different #1 than the global ranking. Ask locally, get a different answer.
The top ten, per AI model
Filled dot = that AI model recommends the brand unprompted when asked "What are the best digital marketing agencies?"
| # | Brand | Score | Named by | ChatGPT | Claude | Gemini | Perplexity | DeepSeek | Grok | Copilot | Google AI |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Ignite Visibility (ignitevisibility.com) | 59.2 | 6/8 | No | Yes | Yes | Yes | No | Yes | Yes | Yes |
| 2 | WebFX (webfx.com) | 47.3 | 5/8 | No | Yes | Yes | Yes | No | Yes | Yes | No |
| 3 | Disruptive Advertising (disruptiveadvertising.com) | 41.4 | 4/8 | Yes | Yes | No | Yes | No | Yes | No | No |
| 4 | SmartSites (smartsites.com) | 32.6 | 3/8 | No | No | Yes | Yes | No | Yes | No | No |
| 5 | Thrive Internet Marketing Agency (thriveagency.com) | 23 | 2/8 | No | No | Yes | No | No | No | Yes | No |
| 6 | KlientBoost (klientboost.com) | 21.3 | 2/8 | No | No | Yes | No | No | No | No | Yes |
| 7 | Funnel Boost Media (funnelboostmedia.net) | 20.7 | 2/8 | No | No | No | No | No | Yes | Yes | No |
| 8 | Victorious (victoriousseo.com) | 20.5 | 2/8 | No | No | Yes | No | No | No | Yes | No |
| 9 | Viral Nation (viralnation.com) | 20.5 | 2/8 | No | No | No | No | No | Yes | No | Yes |
| 10 | VaynerMedia (vaynermedia.com) | 20.3 | 2/8 | Yes | No | No | No | No | Yes | No | No |
AI Recommendation Score 0 to 100; full published formula at citehawk.com/leaderboards/methodology. Snapshot bbf5b6eb, captured 2026-07-02, immutable.
Cross a border, get a different shortlist
| Asked from | #1 in that market | vs global |
|---|---|---|
| United States | WebFX | Different #1 |
| United Kingdom | Hallam | Different #1 |
| Australia | Sparro | Different #1 |
| Canada | Major Tom | Different #1 |
Regional editions are published only where the local ranking meaningfully differs from global. Visibility earned in one market does not automatically transfer.
What to do with this record
If you are on the list
Your position is per-model, and the dot grid shows exactly where it leaks. Each empty dot is one AI model answering your buyers without you, and each has specific causes: the sources that AI model leans on do not carry you, or describe you wrongly. Note that positions held on the two-model minimum are the most fragile on the board: one AI model changing its mind removes the ranking entirely. Claiming your brand on the Index is free, never affects position, and adds movement alerts the moment any AI model changes its answer.
If you are not on the list
The 8 AI models did not name you when buyers asked. That is an invisibility problem with a known playbook: verify what each AI model currently says about you, correct the wrong facts at their sources, and win presence on the sources these answers cite. Pull the receipts on the live category page and list who gets cited alongside the ranked brands; that list is the work.
The cadence that matters
This record re-freezes monthly, and the underlying answers are re-measured weekly on the platform. Rankings move as models retrain and retrieval refreshes, which is why a position is a feed to watch, not a trophy to file.
HVAC Companies
When buyers ask the 8 leading AI models for HVAC companies, these are the names they hear. Measured from the question a real buyer asks, captured verbatim, frozen as the July 2026 record.
- Carrier holds #1 this edition, named by 7 of 8 AI models. ChatGPT does not name it at all.
- 3/10 of the top ten are named by a majority of AI models. 2 hold a ranking on the two-model minimum: fragile positions.
- 3/4 regional editions crown a different #1 than the global ranking. Ask locally, get a different answer.
The top ten, per AI model
Filled dot = that AI model recommends the brand unprompted when asked "What are the best HVAC companies?"
| # | Brand | Score | Named by | ChatGPT | Claude | Gemini | Perplexity | DeepSeek | Grok | Copilot | Google AI |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Carrier (carrier.com) | 63 | 7/8 | No | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 2 | Lennox (lennox.com) | 58.5 | 7/8 | No | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 3 | Trane (trane.com) | 58.2 | 6/8 | No | Yes | Yes | Yes | No | Yes | Yes | Yes |
| 4 | Rheem (rheem.com) | 35.2 | 4/8 | No | No | Yes | Yes | No | Yes | Yes | No |
| 5 | Goodman (goodmanmfg.com) | 34.8 | 4/8 | No | No | Yes | Yes | No | Yes | Yes | No |
| 6 | American Standard (americanstandard.com) | 29.1 | 3/8 | No | No | Yes | No | Yes | Yes | No | No |
| 7 | Bryant (bryant.com) | 26.9 | 3/8 | No | No | Yes | Yes | Yes | No | No | No |
| 8 | Daikin (daikin.com) | 26.6 | 3/8 | No | No | Yes | Yes | No | Yes | No | No |
| 9 | York (york.com) | 18.2 | 2/8 | No | No | Yes | No | No | No | Yes | No |
| 10 | Amana (amana-hvac.com) | 18 | 2/8 | No | No | Yes | No | No | Yes | No | No |
AI Recommendation Score 0 to 100; full published formula at citehawk.com/leaderboards/methodology. Snapshot 18bdeaca, captured 2026-07-22, immutable.
Cross a border, get a different shortlist
| Asked from | #1 in that market | vs global |
|---|---|---|
| United States | Trane | Different #1 |
| United Kingdom | Daikin | Different #1 |
| Australia | Daikin | Different #1 |
| Canada | Carrier | Matches global |
Regional editions are published only where the local ranking meaningfully differs from global. Visibility earned in one market does not automatically transfer.
What to do with this record
If you are on the list
Your position is per-model, and the dot grid shows exactly where it leaks. Each empty dot is one AI model answering your buyers without you, and each has specific causes: the sources that AI model leans on do not carry you, or describe you wrongly. Note that positions held on the two-model minimum are the most fragile on the board: one AI model changing its mind removes the ranking entirely. Claiming your brand on the Index is free, never affects position, and adds movement alerts the moment any AI model changes its answer.
If you are not on the list
The 8 AI models did not name you when buyers asked. That is an invisibility problem with a known playbook: verify what each AI model currently says about you, correct the wrong facts at their sources, and win presence on the sources these answers cite. Pull the receipts on the live category page and list who gets cited alongside the ranked brands; that list is the work.
The cadence that matters
This record re-freezes monthly, and the underlying answers are re-measured weekly on the platform. Rankings move as models retrain and retrieval refreshes, which is why a position is a feed to watch, not a trophy to file.
Law Firms
When buyers ask the 8 leading AI models for law firms, these are the names they hear. Measured from the question a real buyer asks, captured verbatim, frozen as the July 2026 record.
- Kirkland & Ellis holds #1 this edition, named by 6 of 8 AI models. Claude and Google AI do not name it at all.
- 3/10 of the top ten are named by a majority of AI models. 4 hold a ranking on the two-model minimum: fragile positions.
- 4/4 regional editions crown a different #1 than the global ranking. Ask locally, get a different answer.
The top ten, per AI model
Filled dot = that AI model recommends the brand unprompted when asked "What are the best law firms?"
| # | Brand | Score | Named by | ChatGPT | Claude | Gemini | Perplexity | DeepSeek | Grok | Copilot | Google AI |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Kirkland & Ellis (kirkland.com) | 54.9 | 6/8 | Yes | No | Yes | Yes | Yes | Yes | Yes | No |
| 2 | Latham & Watkins (lw.com) | 46.1 | 5/8 | No | No | Yes | Yes | Yes | Yes | Yes | No |
| 3 | Skadden, Arps, Slate, Meagher & Flom (skadden.com) | 45.5 | 5/8 | No | No | Yes | Yes | Yes | Yes | Yes | No |
| 4 | Cravath, Swaine & Moore (cravath.com) | 37.5 | 4/8 | No | No | No | Yes | Yes | Yes | Yes | No |
| 5 | Gibson Dunn (gibsondunn.com) | 34.5 | 4/8 | No | No | Yes | Yes | Yes | No | Yes | No |
| 6 | Sullivan & Cromwell (sullcrom.com) | 27.9 | 3/8 | No | No | Yes | No | Yes | No | Yes | No |
| 7 | Wachtell, Lipton, Rosen & Katz (wlrk.com) | 23.9 | 2/8 | No | No | No | No | No | Yes | Yes | No |
| 8 | Davis Polk & Wardwell (davispolk.com) | 19.4 | 2/8 | No | No | Yes | No | No | No | Yes | No |
| 9 | Vault (vault.com) | 18.7 | 2/8 | No | No | Yes | No | No | Yes | No | No |
| 10 | Simpson Thacher & Bartlett (stblaw.com) | 18.2 | 2/8 | No | No | Yes | No | No | No | Yes | No |
AI Recommendation Score 0 to 100; full published formula at citehawk.com/leaderboards/methodology. Snapshot 86818313, captured 2026-07-22, immutable.
Cross a border, get a different shortlist
| Asked from | #1 in that market | vs global |
|---|---|---|
| United States | Cravath, Swaine & Moore | Different #1 |
| United Kingdom | Linklaters | Different #1 |
| Australia | Allens | Different #1 |
| Canada | Blake, Cassels & Graydon LLP | Different #1 |
Regional editions are published only where the local ranking meaningfully differs from global. Visibility earned in one market does not automatically transfer.
What to do with this record
If you are on the list
Your position is per-model, and the dot grid shows exactly where it leaks. Each empty dot is one AI model answering your buyers without you, and each has specific causes: the sources that AI model leans on do not carry you, or describe you wrongly. Note that positions held on the two-model minimum are the most fragile on the board: one AI model changing its mind removes the ranking entirely. Claiming your brand on the Index is free, never affects position, and adds movement alerts the moment any AI model changes its answer.
If you are not on the list
The 8 AI models did not name you when buyers asked. That is an invisibility problem with a known playbook: verify what each AI model currently says about you, correct the wrong facts at their sources, and win presence on the sources these answers cite. Pull the receipts on the live category page and list who gets cited alongside the ranked brands; that list is the work.
The cadence that matters
This record re-freezes monthly, and the underlying answers are re-measured weekly on the platform. Rankings move as models retrain and retrieval refreshes, which is why a position is a feed to watch, not a trophy to file.
PR Agencies
When buyers ask the 8 leading AI models for PR agencies, these are the names they hear. Measured from the question a real buyer asks, captured verbatim, frozen as the July 2026 record.
- Edelman holds #1 this edition, named by 8 of 8 AI models (unanimous).
- 2/10 of the top ten are named by a majority of AI models. 1 hold a ranking on the two-model minimum: fragile positions.
- 3/4 regional editions crown a different #1 than the global ranking. Ask locally, get a different answer.
The top ten, per AI model
Filled dot = that AI model recommends the brand unprompted when asked "What are the best PR agencies?"
| # | Brand | Score | Named by | ChatGPT | Claude | Gemini | Perplexity | DeepSeek | Grok | Copilot | Google AI |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Edelman (edelman.com) | 64 | 8/8 | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| 2 | Weber Shandwick (webershandwick.com) | 42.1 | 5/8 | Yes | Yes | Yes | No | No | No | Yes | Yes |
| 3 | FleishmanHillard (fleishmanhillard.com) | 31.8 | 4/8 | Yes | No | Yes | No | Yes | No | Yes | No |
| 4 | Real Chemistry (realchemistry.com) | 31.1 | 4/8 | No | No | Yes | Yes | No | Yes | No | Yes |
| 5 | Finn Partners (finnpartners.com) | 30.5 | 4/8 | No | No | No | Yes | Yes | No | Yes | Yes |
| 6 | Unilever (unilever.com) | 29.1 | 2/8 | No | Yes | No | No | Yes | No | No | No |
| 7 | Ogilvy (ogilvy.com) | 24.4 | 3/8 | No | No | Yes | Yes | Yes | No | No | No |
| 8 | Microsoft (microsoft.com) | 24.2 | 3/8 | No | Yes | Yes | No | Yes | No | No | No |
| 9 | Ketchum (ketchum.com) | 24 | 3/8 | Yes | No | Yes | No | Yes | No | No | No |
| 10 | BCW (bcw-global.com) | 23.8 | 3/8 | No | No | No | Yes | Yes | No | Yes | No |
AI Recommendation Score 0 to 100; full published formula at citehawk.com/leaderboards/methodology. Snapshot 2e83e179, captured 2026-07-02, immutable.
Cross a border, get a different shortlist
| Asked from | #1 in that market | vs global |
|---|---|---|
| United States | Edelman | Matches global |
| United Kingdom | Edelman UK | Different #1 |
| Australia | Thrive PR + Communications | Different #1 |
| Canada | NATIONAL Public Relations | Different #1 |
Regional editions are published only where the local ranking meaningfully differs from global. Visibility earned in one market does not automatically transfer.
What to do with this record
If you are on the list
Your position is per-model, and the dot grid shows exactly where it leaks. Each empty dot is one AI model answering your buyers without you, and each has specific causes: the sources that AI model leans on do not carry you, or describe you wrongly. Note that positions held on the two-model minimum are the most fragile on the board: one AI model changing its mind removes the ranking entirely. Claiming your brand on the Index is free, never affects position, and adds movement alerts the moment any AI model changes its answer.
If you are not on the list
The 8 AI models did not name you when buyers asked. That is an invisibility problem with a known playbook: verify what each AI model currently says about you, correct the wrong facts at their sources, and win presence on the sources these answers cite. Pull the receipts on the live category page and list who gets cited alongside the ranked brands; that list is the work.
The cadence that matters
This record re-freezes monthly, and the underlying answers are re-measured weekly on the platform. Rankings move as models retrain and retrieval refreshes, which is why a position is a feed to watch, not a trophy to file.
Roofing Companies
When buyers ask the 8 leading AI models for roofing companies, these are the names they hear. Measured from the question a real buyer asks, captured verbatim, frozen as the July 2026 record.
- GAF holds #1 this edition, named by 5 of 8 AI models. ChatGPT and DeepSeek and Google AI do not name it at all.
- 2/10 of the top ten are named by a majority of AI models. 3 hold a ranking on the two-model minimum: fragile positions.
- 3/4 regional editions crown a different #1 than the global ranking. Ask locally, get a different answer.
The top ten, per AI model
Filled dot = that AI model recommends the brand unprompted when asked "What are the best roofing companies?"
| # | Brand | Score | Named by | ChatGPT | Claude | Gemini | Perplexity | DeepSeek | Grok | Copilot | Google AI |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | GAF (gaf.com) | 46.7 | 5/8 | No | Yes | Yes | Yes | No | Yes | Yes | No |
| 2 | Owens Corning (owenscorning.com) | 45.4 | 5/8 | No | Yes | Yes | Yes | No | Yes | Yes | No |
| 3 | CertainTeed (certainteed.com) | 37.6 | 4/8 | No | Yes | Yes | No | No | Yes | Yes | No |
| 4 | Malarkey (malarkeyroofing.com) | 35.9 | 4/8 | No | Yes | Yes | No | No | Yes | Yes | No |
| 5 | Power Home Remodeling (powerhomeremodeling.com) | 32.7 | 3/8 | No | No | No | Yes | Yes | No | Yes | No |
| 6 | IKO (iko.com) | 28.2 | 3/8 | No | Yes | Yes | No | No | Yes | No | No |
| 7 | Atlas (atlasroofing.com) | 28.2 | 3/8 | No | Yes | Yes | No | No | Yes | No | No |
| 8 | DaBella (dabella.us) | 20.1 | 2/8 | No | No | No | Yes | No | No | Yes | No |
| 9 | Tamko (tamko.com) | 19.5 | 2/8 | No | Yes | Yes | No | No | No | No | No |
| 10 | Bone Dry Roofing (bonedry.com) | 19.3 | 2/8 | No | No | No | Yes | Yes | No | No | No |
AI Recommendation Score 0 to 100; full published formula at citehawk.com/leaderboards/methodology. Snapshot f41fb210, captured 2026-07-22, immutable.
Cross a border, get a different shortlist
| Asked from | #1 in that market | vs global |
|---|---|---|
| United States | GAF | Matches global |
| United Kingdom | Marley | Different #1 |
| Australia | Colorbond Steel | Different #1 |
| Canada | IKO | Different #1 |
Regional editions are published only where the local ranking meaningfully differs from global. Visibility earned in one market does not automatically transfer.
What to do with this record
If you are on the list
Your position is per-model, and the dot grid shows exactly where it leaks. Each empty dot is one AI model answering your buyers without you, and each has specific causes: the sources that AI model leans on do not carry you, or describe you wrongly. Note that positions held on the two-model minimum are the most fragile on the board: one AI model changing its mind removes the ranking entirely. Claiming your brand on the Index is free, never affects position, and adds movement alerts the moment any AI model changes its answer.
If you are not on the list
The 8 AI models did not name you when buyers asked. That is an invisibility problem with a known playbook: verify what each AI model currently says about you, correct the wrong facts at their sources, and win presence on the sources these answers cite. Pull the receipts on the live category page and list who gets cited alongside the ranked brands; that list is the work.
The cadence that matters
This record re-freezes monthly, and the underlying answers are re-measured weekly on the platform. Rankings move as models retrain and retrieval refreshes, which is why a position is a feed to watch, not a trophy to file.
SEO Companies
When buyers ask the 8 leading AI models for SEO companies, these are the names they hear. Measured from the question a real buyer asks, captured verbatim, frozen as the July 2026 record.
- WebFX holds #1 this edition, named by 7 of 8 AI models. Google AI does not name it at all.
- 4/10 of the top ten are named by a majority of AI models. 3 hold a ranking on the two-model minimum: fragile positions.
- 3/4 regional editions crown a different #1 than the global ranking. Ask locally, get a different answer.
The top ten, per AI model
Filled dot = that AI model recommends the brand unprompted when asked "What are the best SEO companies?"
| # | Brand | Score | Named by | ChatGPT | Claude | Gemini | Perplexity | DeepSeek | Grok | Copilot | Google AI |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | WebFX (webfx.com) | 52.9 | 7/8 | Yes | Yes | Yes | Yes | Yes | Yes | Yes | No |
| 2 | Ignite Visibility (ignitevisibility.com) | 48.7 | 6/8 | Yes | Yes | Yes | Yes | No | Yes | Yes | No |
| 3 | Searchbloom (searchbloom.com) | 39.7 | 5/8 | No | Yes | Yes | Yes | No | Yes | Yes | No |
| 4 | SmartSites (smartsites.com) | 38.4 | 5/8 | No | No | Yes | Yes | No | Yes | Yes | Yes |
| 5 | Victorious (victoriousseo.com) | 32.1 | 4/8 | Yes | No | Yes | No | No | Yes | No | Yes |
| 6 | First Page Sage (firstpagesage.com) | 29.6 | 2/8 | No | No | Yes | No | No | No | No | Yes |
| 7 | Intero Digital (interodigital.com) | 24.6 | 2/8 | No | No | No | Yes | No | Yes | No | No |
| 8 | Siege Media (siegemedia.com) | 24.3 | 3/8 | No | No | No | No | Yes | Yes | No | Yes |
| 9 | NP Digital (npdigital.com) | 23.6 | 3/8 | No | No | Yes | No | No | No | Yes | Yes |
| 10 | Victorious (victorious.com) | 17.3 | 2/8 | No | No | No | No | Yes | No | Yes | No |
AI Recommendation Score 0 to 100; full published formula at citehawk.com/leaderboards/methodology. Snapshot 9cb61724, captured 2026-07-02, immutable.
Cross a border, get a different shortlist
| Asked from | #1 in that market | vs global |
|---|---|---|
| United States | WebFX | Matches global |
| United Kingdom | The SEO Works | Different #1 |
| Australia | StudioHawk | Different #1 |
| Canada | Search Engine People | Different #1 |
Regional editions are published only where the local ranking meaningfully differs from global. Visibility earned in one market does not automatically transfer.
What to do with this record
If you are on the list
Your position is per-model, and the dot grid shows exactly where it leaks. Each empty dot is one AI model answering your buyers without you, and each has specific causes: the sources that AI model leans on do not carry you, or describe you wrongly. Note that positions held on the two-model minimum are the most fragile on the board: one AI model changing its mind removes the ranking entirely. Claiming your brand on the Index is free, never affects position, and adds movement alerts the moment any AI model changes its answer.
If you are not on the list
The 8 AI models did not name you when buyers asked. That is an invisibility problem with a known playbook: verify what each AI model currently says about you, correct the wrong facts at their sources, and win presence on the sources these answers cite. Pull the receipts on the live category page and list who gets cited alongside the ranked brands; that list is the work.
The cadence that matters
This record re-freezes monthly, and the underlying answers are re-measured weekly on the platform. Rankings move as models retrain and retrieval refreshes, which is why a position is a feed to watch, not a trophy to file.
Social Media Agencies
When buyers ask the 8 leading AI models for social media marketing agencies, these are the names they hear. Measured from the question a real buyer asks, captured verbatim, frozen as the July 2026 record.
- LYFE Marketing holds #1 this edition, named by 6 of 8 AI models. DeepSeek and Grok do not name it at all.
- 3/10 of the top ten are named by a majority of AI models. 5 hold a ranking on the two-model minimum: fragile positions.
- 3/4 regional editions crown a different #1 than the global ranking. Ask locally, get a different answer.
The top ten, per AI model
Filled dot = that AI model recommends the brand unprompted when asked "What are the best social media marketing agencies?"
| # | Brand | Score | Named by | ChatGPT | Claude | Gemini | Perplexity | DeepSeek | Grok | Copilot | Google AI |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | LYFE Marketing (lyfemarketing.com) | 59.3 | 6/8 | Yes | Yes | Yes | Yes | No | No | Yes | Yes |
| 2 | Sociallyin (sociallyin.com) | 48.7 | 5/8 | No | Yes | Yes | Yes | No | No | Yes | Yes |
| 3 | Viral Nation (viralnation.com) | 46 | 5/8 | No | No | Yes | Yes | No | Yes | Yes | Yes |
| 4 | Disruptive Advertising (disruptiveadvertising.com) | 29.2 | 3/8 | No | No | Yes | Yes | No | No | Yes | No |
| 5 | Fresh Content Society (freshcontentsociety.com) | 28.1 | 3/8 | No | No | Yes | Yes | No | Yes | No | No |
| 6 | Thrive Internet Marketing Agency (thriveagency.com) | 22.5 | 2/8 | No | Yes | Yes | No | No | No | No | No |
| 7 | Sculpt (sculpt.com) | 20.5 | 2/8 | No | No | Yes | No | No | No | Yes | No |
| 8 | Hootsuite (hootsuite.com) | 20 | 2/8 | No | No | No | No | Yes | No | No | Yes |
| 9 | We Are Social (wearesocial.com) | 19.8 | 2/8 | No | No | No | No | No | Yes | Yes | No |
| 10 | The Social Shepherd (thesocialshepherd.com) | 19.8 | 2/8 | Yes | No | Yes | No | No | No | No | No |
AI Recommendation Score 0 to 100; full published formula at citehawk.com/leaderboards/methodology. Snapshot a05c0740, captured 2026-07-02, immutable.
Cross a border, get a different shortlist
| Asked from | #1 in that market | vs global |
|---|---|---|
| United States | LYFE Marketing | Matches global |
| United Kingdom | We Are Social | Different #1 |
| Australia | Megaphone Marketing | Different #1 |
| Canada | ICUC Social | Different #1 |
Regional editions are published only where the local ranking meaningfully differs from global. Visibility earned in one market does not automatically transfer.
What to do with this record
If you are on the list
Your position is per-model, and the dot grid shows exactly where it leaks. Each empty dot is one AI model answering your buyers without you, and each has specific causes: the sources that AI model leans on do not carry you, or describe you wrongly. Note that positions held on the two-model minimum are the most fragile on the board: one AI model changing its mind removes the ranking entirely. Claiming your brand on the Index is free, never affects position, and adds movement alerts the moment any AI model changes its answer.
If you are not on the list
The 8 AI models did not name you when buyers asked. That is an invisibility problem with a known playbook: verify what each AI model currently says about you, correct the wrong facts at their sources, and win presence on the sources these answers cite. Pull the receipts on the live category page and list who gets cited alongside the ranked brands; that list is the work.
The cadence that matters
This record re-freezes monthly, and the underlying answers are re-measured weekly on the platform. Rankings move as models retrain and retrieval refreshes, which is why a position is a feed to watch, not a trophy to file.
Staffing Agencies
When buyers ask the 8 leading AI models for staffing agencies, these are the names they hear. Measured from the question a real buyer asks, captured verbatim, frozen as the July 2026 record.
- Robert Half holds #1 this edition, named by 7 of 8 AI models. Google AI does not name it at all.
- 7/10 of the top ten are named by a majority of AI models. 1 hold a ranking on the two-model minimum: fragile positions.
- 3/4 regional editions crown a different #1 than the global ranking. Ask locally, get a different answer.
The top ten, per AI model
Filled dot = that AI model recommends the brand unprompted when asked "What are the best staffing agencies?"
| # | Brand | Score | Named by | ChatGPT | Claude | Gemini | Perplexity | DeepSeek | Grok | Copilot | Google AI |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Robert Half (roberthalf.com) | 55.5 | 7/8 | Yes | Yes | Yes | Yes | Yes | Yes | Yes | No |
| 2 | Randstad (randstad.com) | 54.3 | 7/8 | Yes | Yes | Yes | Yes | Yes | Yes | Yes | No |
| 3 | Adecco (adecco.com) | 49.6 | 6/8 | Yes | Yes | Yes | Yes | No | Yes | Yes | No |
| 4 | Kelly Services (kellyservices.com) | 47.5 | 6/8 | No | Yes | Yes | Yes | Yes | Yes | Yes | No |
| 5 | Aerotek (aerotek.com) | 46.6 | 6/8 | No | No | Yes | Yes | Yes | Yes | Yes | Yes |
| 6 | Express Employment Professionals (expresspros.com) | 39 | 5/8 | No | No | Yes | Yes | No | Yes | Yes | Yes |
| 7 | Insight Global (insightglobal.com) | 38.7 | 5/8 | No | Yes | Yes | Yes | No | Yes | Yes | No |
| 8 | Manpower (manpowergroup.com) | 27.6 | 3/8 | No | No | Yes | Yes | No | Yes | No | No |
| 9 | TEKsystems (teksystems.com) | 23.6 | 3/8 | No | No | Yes | Yes | Yes | No | No | No |
| 10 | Manpower (manpower.com) | 18.9 | 2/8 | No | Yes | No | No | No | No | Yes | No |
AI Recommendation Score 0 to 100; full published formula at citehawk.com/leaderboards/methodology. Snapshot 1cd3b466, captured 2026-07-22, immutable.
Cross a border, get a different shortlist
| Asked from | #1 in that market | vs global |
|---|---|---|
| United States | Robert Half | Matches global |
| United Kingdom | Reed | Different #1 |
| Australia | Hays | Different #1 |
| Canada | Randstad Canada | Different #1 |
Regional editions are published only where the local ranking meaningfully differs from global. Visibility earned in one market does not automatically transfer.
What to do with this record
If you are on the list
Your position is per-model, and the dot grid shows exactly where it leaks. Each empty dot is one AI model answering your buyers without you, and each has specific causes: the sources that AI model leans on do not carry you, or describe you wrongly. Note that positions held on the two-model minimum are the most fragile on the board: one AI model changing its mind removes the ranking entirely. Claiming your brand on the Index is free, never affects position, and adds movement alerts the moment any AI model changes its answer.
If you are not on the list
The 8 AI models did not name you when buyers asked. That is an invisibility problem with a known playbook: verify what each AI model currently says about you, correct the wrong facts at their sources, and win presence on the sources these answers cite. Pull the receipts on the live category page and list who gets cited alongside the ranked brands; that list is the work.
The cadence that matters
This record re-freezes monthly, and the underlying answers are re-measured weekly on the platform. Rankings move as models retrain and retrieval refreshes, which is why a position is a feed to watch, not a trophy to file.
Web Design Companies
When buyers ask the 8 leading AI models for web design companies, these are the names they hear. Measured from the question a real buyer asks, captured verbatim, frozen as the July 2026 record.
- Clay holds #1 this edition, named by 6 of 8 AI models. ChatGPT and Copilot do not name it at all.
- 1/10 of the top ten are named by a majority of AI models. 7 hold a ranking on the two-model minimum: fragile positions.
- 3/4 regional editions crown a different #1 than the global ranking. Ask locally, get a different answer.
The top ten, per AI model
Filled dot = that AI model recommends the brand unprompted when asked "What are the best web design companies?"
| # | Brand | Score | Named by | ChatGPT | Claude | Gemini | Perplexity | DeepSeek | Grok | Copilot | Google AI |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Clay (clay.com) | 57.7 | 6/8 | No | Yes | Yes | Yes | Yes | Yes | No | Yes |
| 2 | Lounge Lizard (loungelizard.com) | 32 | 3/8 | No | Yes | Yes | No | No | Yes | No | No |
| 3 | Metalab (metalab.co) | 29 | 3/8 | No | Yes | No | No | No | Yes | No | Yes |
| 4 | Big Drop Inc (bigdropinc.com) | 25.5 | 2/8 | No | No | Yes | Yes | No | No | No | No |
| 5 | Dotlogics (dotlogics.com) | 25.5 | 2/8 | No | No | No | Yes | No | Yes | No | No |
| 6 | Fantasy (fantasy.co) | 21.7 | 2/8 | No | No | No | No | Yes | No | No | Yes |
| 7 | Thrive Internet Marketing Agency (thriveagency.com) | 21 | 2/8 | No | No | Yes | No | No | No | Yes | No |
| 8 | SPINX Digital (spinxdigital.com) | 21 | 2/8 | No | No | No | Yes | No | No | No | Yes |
| 9 | Bop Design (bopdesign.com) | 20.7 | 2/8 | No | No | No | No | No | Yes | Yes | No |
| 10 | Slack (slack.com) | 20.7 | 2/8 | No | Yes | No | No | No | No | Yes | No |
AI Recommendation Score 0 to 100; full published formula at citehawk.com/leaderboards/methodology. Snapshot 84a63b89, captured 2026-07-02, immutable.
Cross a border, get a different shortlist
| Asked from | #1 in that market | vs global |
|---|---|---|
| United States | Clay | Matches global |
| United Kingdom | CreativeWeb | Different #1 |
| Australia | Chromatix | Different #1 |
| Canada | Azuro Digital | Different #1 |
Regional editions are published only where the local ranking meaningfully differs from global. Visibility earned in one market does not automatically transfer.
What to do with this record
If you are on the list
Your position is per-model, and the dot grid shows exactly where it leaks. Each empty dot is one AI model answering your buyers without you, and each has specific causes: the sources that AI model leans on do not carry you, or describe you wrongly. Note that positions held on the two-model minimum are the most fragile on the board: one AI model changing its mind removes the ranking entirely. Claiming your brand on the Index is free, never affects position, and adds movement alerts the moment any AI model changes its answer.
If you are not on the list
The 8 AI models did not name you when buyers asked. That is an invisibility problem with a known playbook: verify what each AI model currently says about you, correct the wrong facts at their sources, and win presence on the sources these answers cite. Pull the receipts on the live category page and list who gets cited alongside the ranked brands; that list is the work.
The cadence that matters
This record re-freezes monthly, and the underlying answers are re-measured weekly on the platform. Rankings move as models retrain and retrieval refreshes, which is why a position is a feed to watch, not a trophy to file.
Now see it for your brand.
The paper shows what AI recommends. Your free report shows where your own brand stands in those answers.
Free AI visibility report · No credit card · 50 prompts · 8 engines