# WordPress Drops to 41.5% Market Share > [!question] Lingering Questions > - The "None → vibecoding" causal link is an inference from a correlation in one dataset's bucket, not measured attribution. What would actual platform-level causal data (what a site started as vs. switched to) look like, and does anyone plan to build it? > - Is Tock's "new WP sites are still being created" claim net of churn, or just gross creation? BuiltWith's "~1M new WP sites in the past month" doesn't resolve this. > - How much of the "None" bucket is AI output that's genuinely CMS-less versus fingerprint-less-but-CMS-backed (detection artifact rather than real category exit)? > [!tip] Article Ideas > - "How to read WordPress market-share data" — a plain-language explainer on why W3Techs, HTTP Archive, and BuiltWith legitimately disagree (sampling frame, top-N gating, live-site counting vs. traffic-tier counting). > - The Abilities API is WordPress's answer to the AI squeeze and nobody's talking about it — a DevRel content gap this piece surfaces directly. > [!link] Cross Reference > - [[3. Resources/MOCs/Open Source Governance MOC]] — market-share reading as a governance/strategy input, not just a vanity metric > - [[3. Resources/MOCs/AI & the Future of Knowledge Work MOC]] — vibecoding as category-bypass rather than competitor-migration > - [[3. Resources/Permanent Notes/WordPress Market Share Is a Lagging Indicator That Trails Momentum Loss]] — already promoted from this source in a prior pass > - [[3. Resources/Permanent Notes/WordPress's 2026 Losses Went to No-CMS Sites Not Competitors]] — already promoted from this source in a prior pass > - [[3. Resources/Permanent Notes/Perm · Vibecoding bypasses CMS entirely, draining relative share to 'None']] — already promoted from this source in a prior pass > [!quote] Vocabulary > - **Vibecoding** — Josh Koenig's (Pantheon) two-word name for AI-driven, conversational site-building that produces sites with no detectable CMS fingerprint > - **S-curve maturity** — Noel Tock's (Human Made) framing: a mature product's cumulative install-base metric lags the leading indicators of momentum loss ## Bibliographic capsule The Repository (Rae Morey), part one of a three-part series, lines up W3Techs (41.5%), HTTP Archive (32.95%), and BuiltWith (~40.76% of the whole internet) on WordPress CMS market share and finds they disagree on magnitude, peak date, and where the losses are going — then traces the 2026 movement to AI site-building rather than competitor migration. ## Author's thesis WordPress's market-share decline is real across every dataset, but the dominant single-cause narratives (litigation, competitor poaching) don't survive the timeline or the 2026 data — the actual story is a multi-front squeeze, with the 2026 marginal loss going mainly to AI-built sites that never enter the CMS comparison at all. ## Key arguments (≤3) 1. **The litigation-blame narrative fails on timing.** WordPress hit its all-time-high share (43.6%) after the WP Engine conflict began in September 2024; the decline didn't start until Q2 2025 and didn't accelerate until 2026. 2. **The 2026 losses go mainly to "no CMS," not to Shopify or Wix.** W3Techs H1 2026: WordPress −1.5 points, "None" +1.3, Shopify/Wix combined +0.3 — a category last seen reversing a decade of decline. 3. **Market share itself is a lagging indicator that conceals an earlier momentum turn.** HTTP Archive dates WordPress's professional-tier peak to July 2022; Google Trends search interest peaked around 2014 — years before the headline number moved. ## Claim atoms ### WordPress peaked after the litigation started > [!quote] ID:rw_wpshare01 > WordPress literally peaked after the litigation started. Where/why: the litigation-framing refutation. Load-bearing timing argument that breaks the "lawsuit caused the decline" causal claim — peak of 43.6% came in early 2025, after the September 2024 conflict. ### WordPress losses went to no-CMS, not competitors > [!quote] ID:rw_wpshare02 > The majority of what WordPress lost this year hasn't gone to a competitor. It's gone to sites with no detectable CMS at all — at least, according to W3Techs. Where/why: the "where is the share going?" section, H1 2026 W3Techs figures. The central surprising finding; reframes the whole competitor-migration narrative. ### Koenig names the cause: vibecoding > [!quote] ID:rw_wpshare03 > In a follow-up, he named the cause in two words: "It's vibecoding." Where/why: Josh Koenig (Pantheon) on X, after noting the WP→"None" shift is roughly half of Drupal's entire market share. Names the mechanism behind the "None" reversal. ### Market share is a lagging indicator, not the terrain > [!quote] ID:rw_wpshare04 > The 43%, 42, 41, it's a lagging indicator. The map is not the terrain. Where/why: Noel Tock (Human Made), from his WordCamp Asia 2024 thesis. Reframes market share as backward-looking; argues against steering strategy by the headline number. ### AI users start without a CMS by default > [!quote] ID:rw_wpshare05 > People building sites with AI will usually start without a CMS, because they want a website, and that's what they ask the tool. Where/why: Joost de Valk, on the demand-side mechanism feeding the "None" bucket — the first of his three squeeze vectors (AI, ecommerce-to-Shopify, enterprise-certainty). ## Permanent candidates - All three strong claims in this source were already captured by existing Permanents in a prior synthesis pass. Checked and confirmed no gap: - [[3. Resources/Permanent Notes/WordPress Market Share Is a Lagging Indicator That Trails Momentum Loss]] (H4) - [[3. Resources/Permanent Notes/WordPress's 2026 Losses Went to No-CMS Sites Not Competitors]] (H2 + H3 + H5) - [[3. Resources/Permanent Notes/Perm · Vibecoding bypasses CMS entirely, draining relative share to 'None']] (H2 + H3) - No new Permanent note created from this pass — this Literature Note fills the missing bibliographic/atom record for a source that was already mined for claims. ## Processing checklist - [x] Confirmed this doc's `notes` field already contained a full deep-read (Refactor Appendix present) — ported directly - [x] Checked existing Permanents before creating new ones — found all three candidate claims already promoted; linked instead of duplicating - [x] Connect into the relevant MOC — [[3. Resources/MOCs/Open Source Governance MOC]] ## Links - [[3. Resources/MOCs/Open Source Governance MOC]] — market-share measurement as a governance and strategy question - [[3. Resources/MOCs/AI & the Future of Knowledge Work MOC]] — vibecoding as the mechanism behind 2026's category-bypass losses - [[3. Resources/Readwise/Articles/WordPress Drops to 41.5% Market Share, but Three Datasets Tell Different Stories About What’s Going On]] — source document