Methodology
Version for the November 2026 chart (preliminary)
How the chart works
The CMO Chart ranks chief marketing officers by their public personal brand. The score is our opinion, based on the public signals listed on this page. It says nothing about how good anyone is at marketing.
This chart is preliminary
November 2026 is the first chart we have computed, so it is marked Preliminary — November 2026. In practice that means:
- No arrows. There is no previous month to compare with, so nobody is marked as up, down, new or returning. Arrows start with the next chart.
- The scale is set on this month. Each measure is compared with this month's base. That reference is frozen once the chart goes live, so later months are comparable.
- Some history is short. We started collecting in November 2026, with 90 days of history for press and LinkedIn. Authority will cover a full 12 months as data builds up.
This month in numbers
Data through Nov 21, 2026.
- CMOs we track
- 2,056
- Ranked this month
- 946
- Below the activity bar
- 1,110
- Mentions checked for namesakes
- 6,267
- Earned stories confirmed, 12 months
- 1,326
- LinkedIn posts measured, 90 days
- 8,841
- Company CMOs: 340 Enterprise and 1,617 Growth; 96 university marketing chiefs in the Higher ed chart.
- 422 people (21%) had at least one earned story in the last 90 days, and 1,281 (62%) posted on LinkedIn. Most CMOs are quiet in public, which is why the activity bar exists.
- 1,715 candidate mentions turned out to be about a namesake and were dropped.
- Industry charts published this month: Tech, Financial Services, Healthcare & Pharma, Professional Services, Industrials & Manufacturing, CPG, Retail.
The score
Each month every CMO in our base gets one number from 0 to 100, the CMO Chart Score. It is a weighted sum of five measures:
Score = 0.30 × Visibility + 0.20 × Engagement + 0.20 × Authority + 0.15 × AI recall + 0.15 × Consistency
Each measure is put on the same 0–100 scale: we take the logarithm of the raw value, compare it with a reference month and map it onto 0–100 with the normal curve. The reference is frozen for a year, so a 70 in January and a 70 in June mean the same thing. We recalibrate once a year, in January, and publish the change.
We use a scale, not percentiles, so the gap between No. 1 and No. 2 stays visible. If two people are within 0.1 points of each other, they share a place: the chart shows one number with a bracket and the word “tied”, and in text we write T-4. The next place is skipped (3, 4, 4, 6).
Activity bar. A CMO is ranked when, in the window, they have at least one earned press story in which they are the protagonist or are quoted (12 months), or at least 3 own LinkedIn posts (90 days), or a major award or list (12 months). Everyone else is counted as a Silent CMO and not ranked. Everyone below the bar keeps a profile page with their sources: with no public signal there is nothing to rank.
The five measures
| Measure | Weight | Window | In plain words |
|---|---|---|---|
| Visibility | 30% | last 90 days | How much the press wrote about them. An interview or a column counts in full, a quote in a story about the company counts a little, a passing mention or a press release counts zero. Big-company coverage is scaled down so the employer's brand does not carry the person. |
| Engagement | 20% | last 90 days | How people respond to their LinkedIn posts: the median post, both in absolute terms and per 1,000 followers, so a huge account does not win by size alone. |
| Authority | 20% | last 12 months, fading | The longer track record: press over the past year with older pieces fading, plus awards and major lists. |
| AI recall | 15% | quarterly | Whether AI assistants know who they are without searching the web, asked several times per model and graded by a separate check. Wrong answers do not count. |
| Consistency | 15% | last 90 days | Showing up steadily: active weeks on LinkedIn and months with press, on a square-root curve, so posting three times a day does not help. |
What counts as press
- An interview, profile or column where the person is the main subject counts in full.
- A quote inside a story about the company counts at most a fifth of that.
- A passing mention counts zero. Press releases and the company's own newsroom count zero.
- Outlets are in three tiers (T1 = 3 points, T2 = 2, T3 = 1). The tier list is public.
- The same story syndicated to ten sites counts once.
- Coverage of very large companies is scaled down, so the employer's fame does not carry the person.
Data sources
Who we track
2,056 marketing leaders. 458 of them are named by public sources:
- S&P 500 company: 215
- Fortune 1000 company (2026): 121
- U.S. research or large university (Carnegie 2025): 94
- Forbes World's Most Influential CMOs 2026: 35
- Business Insider Most Innovative CMOs 2026: 22
The rest come from our own research into US companies. For every person we confirmed on their public LinkedIn profile that they hold the role today and are the company's most senior marketer. Profiles show a “Named in” line only for public lists.
What we read
- News: Google News results for each person's name with their company, newest first, collected every week.
- Awards and lists: public lists such as Forbes and Business Insider, plus awards found in the news.
- Talks and podcasts: event pages, recordings and coverage found the same way.
- LinkedIn: public posts and their reactions, comments and reposts, plus follower counts. We do not use private data or messages.
- AI assistants: our own test prompts to several models (see AI recall).
Every mention is checked to be the same person, not a namesake, before it counts: we read the page around the name and compare it with the person's role, company and history. If we are not sure, or the page cannot be read, it does not count.
When it updates
- The chart is published once a month, on the first Tuesday, covering the previous calendar month.
- News, awards and talks are collected every week, on Mondays. LinkedIn posts and their engagement are refreshed every week too.
- Job changes are checked once a month: someone who leaves a CMO role leaves the chart.
- AI recall is measured once a quarter, so it changes four times a year.
- About half of the score moves slowly on purpose (authority over 12 months, consistency over 90 days, AI recall quarterly). The chart moves slower than a music chart. That is honest: most CMOs have zero to two press mentions in a month.
Why arrows move
Small changes in rank are mostly noise. One extra quote can move someone at No. 80 by twenty places. So we only draw an arrow when the move is real:
- For each person we estimate the range their rank would likely fall in if we re-ran the month with slightly different data (a 90% range). Lower in the chart, the range is wider.
- If this month's rank is outside last month's range, the move is significant: 3 up or 2 down, with the number of places.
- If it is inside the range, we show =, even if the raw rank changed by a few places. “=” is a real result, not an empty cell.
- Every arrow gets a one-line reason with links: “Interview in an outlet; keynote at a conference; LinkedIn engagement ×1.4”.
Solid bracket: the 90% range of the rank
Square: this month's rank
Arrows: from the second live chart on
Every chart row draws this as a strip: the rank axis runs from worse on the left to No. 1 on the right, the dashed bracket is last month's range, the solid bracket and square are this month's. The chart rows use a log scale, so a range near the top and a range near No. 100 are both readable.
Colour means significance, not good or bad. Orange marks an earned arrow, up or down alike, and new or returning entries. Direction is carried by the shape (filled ▲ up, outlined ▽ down) and the number, so it reads in greyscale and for colour-blind readers.
NEW means a first month in the chart. RE means back after a gap (we look back over the published charts).
Each month we check how many significant moves are explained by a real event (a T1 article, an award, a new job, a standout post). If fewer than 70% are, the method is too noisy and we fix it before publishing.
June (award season) and December (holidays) are unusual months. Charts for those months carry a note.
AI recall
More and more people look up executives by asking an AI assistant. AI recall measures whether assistants know who someone is.
- We ask three leading models about each person with web search turned off: once per model, and four more times when a model seems to know them.
- We do not give the company name in the question, so the model has to know it.
- A separate check grades the answers against the person's real career. A confident wrong answer counts as not knowing, and so does an answer that gets one past job right but invents the current one.
- We also ask “who are the most influential CMOs in [industry]?” and count how often the person is named.
A profile page on this site that the person links to can help assistants learn the right facts. We never sell that as a service.
Who is on the chart
- The most senior marketer of a company: CMO, or a Chief Brand, Growth or Customer Officer who runs marketing. Not heads of sub-brands or divisions.
- Companies with 1,000+ employees, or $250M+ revenue, or Series C funding and beyond. United States first.
- Four main charts: All (Top 100 company CMOs), Enterprise (Top 50; S&P 500, Fortune 500, their subsidiaries and companies with 10,000+ staff), Growth (other companies, Top 50) and Higher ed (US universities, Top 50), plus the B2B and B2C charts.
- B2B and B2C charts and industry charts rank the same score within a group (see B2B and B2C charts). We publish one only when at least 20 people in it are ranked.
- Anyone in the base has a profile page, even outside the Top 100. Anyone can suggest a CMO we missed.
B2B and B2C charts
Business-to-business and consumer marketing build very different public profiles, so besides the filter on every chart there are separate B2B and B2C charts. They use the same score as the main chart and rank people only against others of the same business model.
- How a company is classed. Each company gets one label from its main customers: B2B (sells mainly to businesses: software, industrial goods, professional services), B2C (sells mainly to consumers: retail, consumer goods, media, travel) or B2B2C.
- B2B2C. Companies that sell to both businesses and consumers (payment networks, telecoms, large platforms) will appear in both the B2B and the B2C chart, ranked separately. They are not in either chart this month yet, but they are in the main chart.
- Universities are not in the B2B or B2C charts; they have their own Higher ed chart.
- This month: 699 people ranked in B2B and 154 in B2C.
Keeping it fair
- We use the median post, not the average, so one viral post does not dominate.
- Posting more has diminishing returns (a square-root curve). Posting three times a day does not help.
- Engagement pods, “comment YES below” bait and templated comments are discounted.
- Reposts of company content do not count as posts.
- Paid placements, paid “contributor” lists and press releases get zero or minimal weight.
- Claiming a profile never changes the score. There is no paid placement.
- Weights, windows and registries are public. The specific fraud checks are not, so they keep working.
- Once a quarter we publish how the Top 100 breaks down by gender, industry and company size, so blind spots are visible.
What we do not measure, and known limits
- We do not measure business results or the quality of anyone's marketing.
- We favour English-language press and LinkedIn. People who build their brand elsewhere are under-counted.
- Slow measures mean a great month shows up gradually.
- If LinkedIn data becomes unavailable, we switch to a published fallback (Visibility 45%, Authority 35%, AI recall 20%) and say so on the chart.
- We will check how well the score matches respected annual CMO lists and publish the result here, whatever it is.
Corrections and appeals
- Anyone can report a factual error: a wrong job, a namesake's article, a missed award. Send a correction.
- We reply within 10 working days. The editor decides (with a member of the advisory board once it is in place), and the outcome goes into a public log.
- Our own factual errors are fixed right away. Other changes apply from the next month's chart.
Opting out
Anyone can ask to be removed at any time, no reason needed. We take you off all charts within 30 days and add you to a do-not-track list, so we stop collecting data about you. Request removal.
Scoring version ccs-1.0, November 2026. Every later change to weights, windows or thresholds will be logged here with a version number.