Open source leadership can lean toward one employer without any rule saying so. In my count, Automattic held more WordPress team rep seats than any company every year from 2018 to 2026. Three of the ten projects compared here cap one employer’s share. WordPress has no such cap.
Key takeaways
- The big picture: Automattic led WordPress rep seats every year from 2018 to 2026, with 13 to 26 percent.
- By the numbers: It never held a rep seat on every team. The peak was 8 of 14 teams, in 2023.
- Why it matters: Node.js, OpenJS and W3C write limits on one employer. WordPress, PHP and Drupal don’t.
- Yes, but: Automattic’s sponsored contributors reach 22 teams. Reps are a smaller, separate count.
🧭 What counts as open source leadership here?
If you’ve filed a Five for the Future pledge, you know the page it makes. It counts people, hours and teams. Automattic’s page, for example, reads 93 contributors and 2,134 hours a week across 22 teams.1 However, a pledge page never says who runs those teams.
That’s this post’s question. In how many open source board seats sponsorship can buy, I counted board seats. This post moves down a level, to open source leadership inside teams: reps, chairs, steering committees and maintainers.
In WordPress, the leadership role inside a team is the team rep. Anyone active on a team can stand, and elections happen every year.2 A rep runs meetings, posts updates and raises the team’s issues with the rest of the project. In other words, reps are the team’s voice in the room.
So the question becomes simple to ask, if slow to answer. When companies pay for the work, who holds the rep seats? And does any rule stop one company from holding most of them?
If your company pays your hours, this is about you, too. A rep seat is one of the few roles where a sponsor’s name meets a team’s plans. And if you volunteer, it’s about who you’ll share the table with.
That second question isn’t only a WordPress question. So the post also compares open source leadership rules in nine other projects, from Node.js to Joomla.
📊 WordPress team reps by employer, 2018 to 2026
| Year | Rep seats | Automattic | Other employer | Self, no employer named | Unknown | Teams with reps | Teams with an Automattic rep |
|---|---|---|---|---|---|---|---|
| 2018 | 28 | 6 (21%) | 19 | 0 | 3 | 11 | 4 |
| 2019 | 21 | 3 (14%) | 15 | 0 | 3 | 12 | 3 |
| 2020 | 39 | 8 (21%) | 27 | 0 | 4 | 13 | 6 |
| 2021 | 34 | 6 (18%) | 24 | 3 | 1 | 13 | 5 |
| 2022 | 43 | 7 (16%) | 33 | 2 | 1 | 14 | 5 |
| 2023 | 38 | 10 (26%) | 25 | 2 | 1 | 14 | 8 |
| 2024 | 47 | 8 (17%) | 38 | 1 | 0 | 15 | 6 |
| 2025 | 53 | 7 (13%) | 43 | 2 | 1 | 16 | 5 |
| 2026 | 47 | 6 (13%) | 34 | 4 | 3 | 20 | 5 |
Here’s what the table shows, in four lines.
- Most seats: In my count, Automattic held the most rep seats of any employer every year from 2018 to 2026.
- Its share: That came to 3 to 10 seats a year, or 13 to 26 percent.
- Runners up: No other employer held more than 3 seats in a year.
- Every team: Automattic never held a rep seat on every team. Its high was 8 of 14 teams, in 2023.
“Other employer” isn’t one company, either. In 2025, for example, 43 seats went to dozens of different employers, and none held more than 3. So the spread of open source leadership in WordPress is wide, with one tall column in the middle of it.
🧮 How the count works
The record behind each seat comes from public sources, in this order. Because a seat is one person on one team in one year, a handover can add a person without adding a chair at the table. The Merge Conflicts methods page explains the labels.
- Rep posts: each team’s public announcements on the Make WordPress blogs name its reps.
- Pledge profiles: archived WordPress.org profiles show who sponsored each rep’s hours that year.
- Outside records: term research and public LinkedIn histories fill gaps. People write their own LinkedIn pages, so those rows carry a lower label.
- Public chat: searches of public Making WordPress Slack channels found one more employer.
- My own seats: my 2015 seat counts as self employed, from my own account.
Two things stay out on purpose. First, the table counts seats and never names reps. Second, no direct message or private channel went into any count.
🕰️ Before 2018, the record thins
From 2012 to 2017, the record holds 35 rep seats, and 8 still have no employer on file. Automattic held 2 of 13 in 2012. In 2013, DreamHost led with 2 of 10. With numbers that small, though, a single seat moves a share by 8 points or more. From 2014 to 2017, for instance, the record holds just 5, 1, 2 and 4 seats a year. So the table starts in 2018, when the rep posts get regular.
🗣️ What I was told, and what I saw
I’ve been a Training Team rep more than once: in 2015, and again from 2019 to 2022. In 2019 and 2020, I was told that the aim was an Automattic employee in the rep role on every team. Who said it stays private, because it wasn’t said in public.
I also watched it happen, or start to. I saw it in rep chats across many teams, and closest in Training, Community, Marketing and Docs. Automattic staff joined team after team. I saw it, too, in how work got coordinated between teams. Meanwhile, my own team was struggling. In fact, the Training Team wasn’t functioning then.
As I saw it, the aim ended with Automattic’s 2025 cut. That’s my reading, not a statement anyone made.
🔍 What the record shows beside it
An aim is a claim about intent, so it stays my account. Here’s the public record beside it, element by element.
📜 What the project said about team reps in public
The project and Automattic both wrote about sponsored work and reps in public. Here’s what they said, in order.
- When coordination went public: A March 2021 post gave its purpose first: helping coordinate sponsored work. It then opened a Slack channel “for contributors who are sponsored for 32 hours or more.”5
- What the project asked for: That November, a Marketing Team post relayed the executive director’s wish. She wanted “two to three Team Reps in each Contributor Team,” with “a mix of company sponsored and self sponsored folks.”6
- When it changed: Automattic’s January 2025 post gave its reasons first. It cited “an imbalance in how contributions to WordPress are distributed” and the cost of WP Engine’s lawsuits. Then it cut its pledge to “about 45 hours a week.”7 In May 2025, it announced a return.8
- How it was framed: The January post had a heading, “A Level Playing Field.” Under it, Automattic said it would match pledges “made by WP Engine and other players.”9
- What the pledge page showed: Automattic listed 103 profiles in November 2024 and 86 in February 2025.10
🧾 Where the record and my account meet
So the record matches part of my account. Contributors roughly tripled from 2019 to 2022. Rep seats, however, never reached every team. The public wish, two to three reps per team with a mix of sponsors, leaves room for what I was told. It neither confirms nor contradicts it.
I found no public message stating the aim, after 180 searches of public team channels. Still, unknown isn’t the same as absent.
🧢 Contributors aren’t reps
These are two different counts, and they’re easy to mix up.
Also, a rep doesn’t decide alone. The Core handbook describes reps who come from every corner of the project, with “different levels of familiarity.”12 In practice, the role coordinates more than it commands. Still, reps set agendas, run meetings and speak for their teams. That’s open source leadership, even when it’s quiet.
Put simply, a pledge page counts hours, while a rep roster counts voices. So read the two side by side. The gap between them shows how much of a team’s work, and of its voice, comes from one place.
🧱 Which projects cap one employer’s share of open source leadership?
Some projects write a limit on how many leaders one employer can supply. Most don’t. In open source leadership, a cap like this keeps one payroll from becoming the project’s whole voice. Here’s the record for ten projects, including WordPress.
| Project | Leadership body | Written limit on one employer | Reason given beside the rule |
|---|---|---|---|
| Drupal | Core committers and initiative leads | None found | None found |
| Joomla | Department coordinators | None found in the bylaws13 | None found |
| MariaDB | Stewards and maintainers | None; the 25 percent cap covers the board only14 | The process applies to every maintainer “regardless of employer or affiliation” |
| MySQL | Committers, leads and steering committee | Oracle nominates the first members; the FAQ says Oracle holds the majority, but the listed roster doesn’t show one1516 | None stated |
| Node.js | Technical Steering Committee | No more than one fourth of voting members | To ensure “adequate representation” of diverse interests |
| OpenJS | Cross Project Council | No more than one fourth of voting members | None beside the rule |
| PHP | RFC voters and release managers | None found | None found |
| TYPO3 | Team leads | None found for teams; the board allows one person per company.17 Members voted in October 2026 on rules that regroup teams into Units18 | None found |
| W3C | Elected TAG and Advisory Board seats; group chairs | One elected seat per primary affiliation; a chair can be replaced if a new employer would be “over-represented” | “to ensure that the diversity of W3C Members is represented” |
| WordPress | Team reps | None; each team runs its own yearly election | None found |
So three bodies write a limit: Node.js, OpenJS and W3C.192021 In each case, a change of job can trigger it. The Node.js TSC charter, for example, says the overflow “must be immediately remedied.”22 OpenJS uses almost the same words for its council.23
W3C works a little differently. Its limit covers elected seats, and the Team can also replace a group chair after a job change.24 Meanwhile, MySQL sits at the other end. Oracle’s own governance page says committers and project leads are, at first, “nominated by Oracle.”25
None of these caps names a company. Instead, each counts employers, whoever they turn out to be. That’s what makes a cap neutral: it would bind any employer that grew large.
🔁 What happens when a leader changes jobs
A cap matters most on the day someone switches employers. So each charter says what happens next.
- Node.js: The overflow is fixed “by the removal of voting member status” from members tied to the over-represented company.26
- OpenJS: The council fixes it “by the resignation or removal of one or more voting CPC members.”27
- W3C: On the TAG, the organization “must have returned to having at most one seat” by the next regular election.28
In each case, the person stays welcome on the team. Only their vote or seat moves. That keeps the work and changes the count.
⚖️ Share of the work against share of the votes
A cap counts leaders. However, the work and the votes can split in other ways too.
So a large share of the code doesn’t automatically mean a large share of the say. Then again, a small share of the say doesn’t mean a small share of the influence. MariaDB’s process, for its part, says it applies to every maintainer “regardless of employer or affiliation.”32 That’s a rule about equal treatment, not a cap on numbers.
🪢 Concentration risk, without a villain
Concentration isn’t a charge against anyone. It’s a shape, and shapes have risks.
When one employer supplies many leaders, one decision lands on many teams at once. For instance, Automattic’s rep seats went from 10 in 2023 to 6 by 2026. Its pledge page listed 112 profiles in July 2023 and 86 in February 2025, after the cut. By December 2025, it listed 88.33 These records don’t explain any single departure.
On the other hand, paid time also kept teams running. As I said, Training was struggling in those years. Help from any employer would have mattered.
A cap has costs, too. On a small team, it could keep a willing volunteer out of a lead role just because of where they work. Also, a cap spreads formal authority. It doesn’t staff the work.
That’s the trade. One employer can staff a team quickly. It can also leave quickly, and without a cap, nothing in the rules spreads that risk.
So what would spread it? The projects above offer two tools. First, a cap on one employer’s share of open source leadership, as in Node.js and OpenJS. Second, a stated mix, like the 2021 wish for sponsored and self sponsored reps on every team. WordPress has the second as a public wish, but not the first as a rule.
❓ Check your project’s cap on open source leadership
My ask is simple: check your project’s cap. These questions match the tables.
- Limit: Does any rule cap how many leaders one employer can supply?
- Trigger: What happens when a leader changes jobs?
- Count: Who holds your team’s lead roles today, and who pays for their time?
- Unknowns: How many seats have no employer on record?
- Reason: If there’s a cap, does the rule say why?
You don’t need a cap to ask these. However, if the answers come back “no,” “nothing” and “we don’t know,” you’ve found the gap this post is about. Then bring the count to your next team meeting, and ask whether a written limit would help.
🙋 Frequently asked questions
No. Each team runs its own rep election, and no written rule limits one employer’s share. A 2021 post did state a wish for a mix of sponsored and self sponsored reps.
Not in the public record. Its highest was 8 of 14 teams, in 2023.
Node.js and OpenJS cap one employer at a fourth of voting members. W3C allows one elected seat per affiliation on its TAG and Advisory Board. Those three are the only written caps on open source leadership in this comparison.
Next up, 2.6 asks whether publishing your finances gives members a say.
Every relationship I disclose here
- Pantheon: my employer, a WordPress and Drupal host and a WP Engine competitor.
- FAIR Package Manager: an active member of its Technical Steering Committee.
- GoDaddy: I managed its Five for the Future contributors until July 2026, and it sponsored my Training rep hours in 2022.
- The WPOCC: a founding board member.
- Training Team: former faculty member and rep.
Footnotes (33)
- WordPress.org. “Automattic.” Five for the Future. https://wordpress.org/five-for-the-future/pledge/automattic/. Accessed 2026-11-15. ↩︎
- Make WordPress Core. “Project Organization.” Core Contributor Handbook. https://make.wordpress.org/core/handbook/about/organization/. Accessed 2026-11-15. ↩︎
- WordPress.org. “Courtney Robertson (courane01).” WordPress.org profile, archived 2022. Internet Archive. https://web.archive.org/web/2022/https://profiles.wordpress.org/courane01/. Accessed 2026-11-15. ↩︎
- WordPress.org. “Automattic.” Five for the Future, archived captures of 2019-12-19, 2022-07-01, 2023-07-14, 2024-11-17, 2025-02-12, 2025-07-09 and 2025-12-08 (replace the date in the address to open each). Internet Archive. https://web.archive.org/web/20191219/https://wordpress.org/five-for-the-future/pledge/automattic/. Accessed 2026-11-15. ↩︎
- Make WordPress Updates. “Experiment: Coordinating Sponsored Contributor Teams.” WordPress.org. 2021-03-19. https://make.wordpress.org/updates/2021/03/19/experiment-coordinating-sponsored-contributor-teams/. Accessed 2026-11-15. ↩︎
- Make WordPress Marketing. “2022 team nominations for marketing.” WordPress.org. 2021-11-22. https://make.wordpress.org/marketing/2021/11/22/2022-team-nominations-for-marketing/. Accessed 2026-11-15. ↩︎
- Automattic. “Aligning Automattic’s Sponsored Contributions to WordPress.” Automattic. 2025-01-09. https://automattic.com/2025/01/09/aligning-automattics-sponsored-contributions-to-wordpress/. Accessed 2026-11-15. ↩︎
- Automattic. “Returning to Core.” Automattic. 2025-05-29. https://automattic.com/2025/05/29/returning-to-core/. Accessed 2026-11-15. ↩︎
- Automattic. “Aligning Automattic’s Sponsored Contributions to WordPress.” Automattic. 2025-01-09. https://automattic.com/2025/01/09/aligning-automattics-sponsored-contributions-to-wordpress/. Accessed 2026-11-15. ↩︎
- WordPress.org. “Automattic.” Five for the Future, archived captures of 2019-12-19, 2022-07-01, 2023-07-14, 2024-11-17, 2025-02-12, 2025-07-09 and 2025-12-08 (replace the date in the address to open each). Internet Archive. https://web.archive.org/web/20191219/https://wordpress.org/five-for-the-future/pledge/automattic/. Accessed 2026-11-15. ↩︎
- WordPress.org. “Automattic.” Five for the Future. https://wordpress.org/five-for-the-future/pledge/automattic/. Accessed 2026-11-15. ↩︎
- Make WordPress Core. “Project Organization.” Core Contributor Handbook. https://make.wordpress.org/core/handbook/about/organization/. Accessed 2026-11-15. ↩︎
- Open Source Matters. “By-Laws of Open Source Matters, Inc.” Open Source Matters. 2021-12-13. https://www.opensourcematters.org/images/documents/20211213-OSM-Bylaws.pdf [PDF]. Accessed 2026-11-15. ↩︎
- MariaDB Foundation. “By-laws.” MariaDB.org. https://mariadb.org/bylaws/. Accessed 2026-11-15. ↩︎
- Oracle. “MySQL Community Governance Model.” MySQL. https://dev.mysql.com/community/governance-model/. Accessed 2026-11-15. ↩︎
- Oracle. “MySQL Community Steering Committee.” MySQL. https://dev.mysql.com/community/steering-committee/. Accessed 2026-11-15. ↩︎
- TYPO3 Association. “Statuten der TYPO3 Association.” TYPO3 Policy Guide. https://docs.typo3.org/m/typo3/guide-policy/main/en-us/Association/BylawsAndProceedingsGermanOriginal.html. Accessed 2026-11-15. ↩︎
- Stefan Busemann, TYPO3 Association. “TYPO3 Association Members to Vote on New Governance Structure.” TYPO3 News. 2026-10-15. https://news.typo3.com/article/typo3-association-members-to-vote-on-new-governance-structure. Accessed 2026-11-15. ↩︎
- Node.js. “TSC Charter.” GitHub. https://github.com/nodejs/TSC/blob/main/TSC-Charter.md. Accessed 2026-11-15. ↩︎
- OpenJS Foundation. “CPC Charter.” GitHub. https://github.com/openjs-foundation/cross-project-council/blob/main/CPC-CHARTER.md. Accessed 2026-11-15. ↩︎
- W3C. “W3C Process Document.” W3C. https://www.w3.org/policies/process/. Accessed 2026-11-15. ↩︎
- Node.js. “TSC Charter.” GitHub. https://github.com/nodejs/TSC/blob/main/TSC-Charter.md. Accessed 2026-11-15. ↩︎
- OpenJS Foundation. “CPC Charter.” GitHub. https://github.com/openjs-foundation/cross-project-council/blob/main/CPC-CHARTER.md. Accessed 2026-11-15. ↩︎
- W3C. “W3C Process Document.” W3C. https://www.w3.org/policies/process/. Accessed 2026-11-15. ↩︎
- Oracle. “MySQL Community Governance Model.” MySQL. https://dev.mysql.com/community/governance-model/. Accessed 2026-11-15. ↩︎
- Node.js. “TSC Charter.” GitHub. https://github.com/nodejs/TSC/blob/main/TSC-Charter.md. Accessed 2026-11-15. ↩︎
- OpenJS Foundation. “CPC Charter.” GitHub. https://github.com/openjs-foundation/cross-project-council/blob/main/CPC-CHARTER.md. Accessed 2026-11-15. ↩︎
- W3C. “W3C Process Document.” W3C. https://www.w3.org/policies/process/. Accessed 2026-11-15. ↩︎
- The PHP Foundation. “The PHP Foundation Impact and Transparency Report 2025.” The PHP Foundation. 2026-05-27. https://thephp.foundation/blog/2026/05/27/impact-and-transparency-report-2025/. Accessed 2026-11-15. ↩︎
- MariaDB. “Maintainers.” MariaDB governance repository, GitHub. https://github.com/MariaDB/governance/blob/main/lists/maintainers.md. Accessed 2026-11-15. ↩︎
- Ben Ramsey. “PHP RFC: PHP License Update.” PHP Wiki. 2025-07-10. https://wiki.php.net/rfc/php_license_update. Accessed 2026-11-15. ↩︎
- MariaDB. “Governance.” MariaDB governance repository, GitHub. https://github.com/MariaDB/governance/blob/main/governance.md. Accessed 2026-11-15. ↩︎
- WordPress.org. “Automattic.” Five for the Future, archived captures of 2019-12-19, 2022-07-01, 2023-07-14, 2024-11-17, 2025-02-12, 2025-07-09 and 2025-12-08 (replace the date in the address to open each). Internet Archive. https://web.archive.org/web/20191219/https://wordpress.org/five-for-the-future/pledge/automattic/. Accessed 2026-11-15. ↩︎


Leave a Reply