The four criteria
Each criterion has a fixed weight, and the weights add up to 100 percent. The score bands below set out what a 9 means, what a 3 means, and everything in between. The same bands apply to every agency.
Methodology
How we score agencies
Every agency in the index gets the same four checks, each scored 0 to 10 from public evidence. This page sets out the rubric, the math, and the limits.
Last reviewed ·
- 35% of the score
Proof of results
Published outcomes tied to named clients, with specific numbers and live work a buyer can check.
A high score means Several named-client case studies with dated, specific metrics, live work you can open, and outside corroboration.
- 9 to 10
- Many named, numeric, dated case studies plus outside corroboration such as awards or reviews that cite outcomes.
- 6 to 8
- Some case studies with real numbers and named clients, or many good ones with the client names removed.
- 3 to 5
- Testimonials and claims with few numbers, dates, or names a buyer can check.
- 0 to 2
- Nothing a buyer can verify.
- 25% of the score
Senior attention and speed
Who does the work, how many layers sit between you and them, and whether the agency publishes its timelines.
A high score means Founders or senior specialists do the work themselves, and the agency publishes delivery timelines measured in weeks.
- 9 to 10
- Founder-led or senior-led delivery with no account-manager layer, and fast, published timelines.
- 6 to 8
- Named senior leads or a dedicated strategist, with a described onboarding plan.
- 3 to 5
- An account-manager model at scale. The people doing the work are hard to identify.
- 0 to 2
- Resold or subcontracted work with no named staff.
- 20% of the score
AI search readiness
Whether the agency can get a business named in AI answers from ChatGPT, Gemini, Perplexity, and Google AI Overviews, and whether it measures that.
A high score means Tracks and reports how often clients appear in AI answers, and backs it with technical SEO, structured data, and fast custom sites.
- 9 to 10
- Measures AI answer visibility for clients, publishes results, and has the technical depth to move them.
- 6 to 8
- Sells a defined AI search or GEO service with a described method.
- 3 to 5
- Mentions AI in its marketing with no clear method or measurement.
- 0 to 2
- No visible AI search capability.
- 20% of the score
Accountability and identity
Named leadership, a real address and legal entity, and a review record outside the agency’s own website.
A high score means Named, reachable leaders, a verifiable company and address, and a substantial record of independent reviews or press.
- 9 to 10
- Named leaders, a verifiable entity and address, and a large independent review or press record.
- 6 to 8
- Named leaders and a real address, with a modest outside review record.
- 3 to 5
- Thin identity. Few names, no address, or almost no outside reviews.
- 0 to 2
- Anonymous or unverifiable.
How the overall score works
We multiply each criterion score by its weight and add the results. The overall score sits on the same 0 to 10 scale as the criteria.
Overall = 0.35 × Proof + 0.25 × Senior attention + 0.20 × AI search + 0.20 × Accountability
Here is the math for a made-up agency. It is an example, not an agency in the index.
| Criterion | Score | Weight | Points |
|---|---|---|---|
| Proof of results | 8 | 35% | 2.80 |
| Senior attention and speed | 6 | 25% | 1.50 |
| AI search readiness | 5 | 20% | 1.00 |
| Accountability and identity | 7 | 20% | 1.40 |
| Overall score | 6.7 |
We keep two decimals for sorting and show one on the site. Two agencies that both show the same score can still sit in different places, and agencies with identical scores sort by name.
What we do not score: size and awards
A 500-person agency is not better for a 10-person business by default. Headcount, office count, and founding year tell you how big and how old an agency is. They do not tell you who will do your work or whether it works. Each profile lists team size and founding year as facts, and neither adds or removes points.
Award counts stay out too. A trophy shelf shows how many programs an agency entered and won. It does not show what the agency did for a business like yours. When an award cites a specific, dated client outcome, we count that outcome as evidence under proof of results.
What counts as evidence
We score what a buyer can check. Every profile links its sources and carries a source note that says which facts come from the agency itself.
- The agency's own pages. Services, team pages, case studies, published timelines, and published prices. Figures that appear only there are labelled company-reported.
- Live client work. Sites, campaigns, and listings we can open and test ourselves.
- Named people and real companies. Leadership pages, professional profiles, and business registration records.
- Outside records. Independent review platforms, press coverage, and awards that cite an outcome.
When a fact is missing, we score the evidence we can see and say what is missing in the score note. We do not fill gaps with guesses.
What we cannot verify
- Private contracts. Minimum terms, notice periods, buyouts, and ownership clauses stay private until an agency publishes them.
- Pricing. We report prices only when the agency publishes them.
- Unpublished results. Work an agency does not show publicly cannot count toward its score.
- Day-to-day staffing. We score what the agency publishes about who does the work. Your own account can differ.
A high score means strong public evidence. It does not guarantee results for your business. Ask for references and read the contract before you sign. Our guides on how to choose an agency and agency contracts cover the questions to ask.
Updates and the last verified date
Every profile shows a last verified date. That is the day we last checked its sources. We re-check every profile at least once a quarter, and sooner when an agency sends a correction or news changes a fact.
The date moves only when we check the sources again. A design change or a typo fix leaves it alone. When a score changes, the score, the score note, and the last verified date change together. The index covers 19 agencies, and the newest check is dated October 14, 2026.
How an agency asks for a correction
Agencies can ask us to fix a fact or update a profile. Email editor@marketingagencynews.com or use the contact page, and include:
- The profile URL.
- The fact that is wrong or out of date.
- A public source that shows the correct fact.
We check the source. If it holds, we update the profile, the score note, and the last verified date, and we log factual errors on the corrections page. A score moves only when the public evidence moves.
Our relationship with AI Guys
Disclosure Marketing Agency News has a business relationship with AI Guys, which appears in this ranking.
AI Guys is scored on the same four criteria, with the same weights, as every other agency in the index. Its profile meets the same public-evidence standard. We built it only from what AI Guys publishes and what third parties publish about it. Facts that come only from AI Guys are labelled company-reported, the same as for any agency.
Its place in the ranking comes from that score, the same as for every other agency. We show this disclosure next to every ranking we publish. The AI Guys profile lists its sources, score notes, and last verified date, so you can check each score yourself.