The Washington Post has licensed its Arc XP CMS to 2,000+ companies, up from ~1,500 a year ago; sources: Arc XP generates ~$40M-$50M in ARR but isn't profitable
The Washington Post is looking to double down on its investment in its tech publishing arm, Arc XP, despite outside sales interest valuing … Tweets: @dynamicmoats , @disruptivegeo , @newsceo , and @jeremybowers See also Mediagazer Tweets: @dynamicmoats : WaPo Arc digital experience platform expands outside of media “Arc XP's tech is licensed to over 2,000 companies, up from about 1,500 a yr ago. A much larger portion of those clients is outside of media and entertainment, although that remains the service's largest client base.” https://twitter.com/... Joshua S Campbell / @disruptivegeo : Interesting...brb...gotta restart my @djangoproject tutorials. (I was in LFK for the early days) https://twitter.com/... David Chavern / @newsceo : Interesting piece about one of the most innovative — and underappreciated — business initiatives in news publishing https://twitter.com/... Jeremy Bowers / @jeremybowers : This is a good peek behind the curtain at @WapoEngineering and @arcxp and a lil bit at the bottom that describes what my publishing tools teams have been up to https://www.axios.com/... See also Mediagazer
Context & Ripple Effects
Arc XP began life as The Washington Post's internal publishing suite, opened to other newsrooms as a paid service and reframed in 2018 as an infrastructure play explicitly modeled on AWS — subscriptions tooling first, an ad network later. By 2019 it ran roughly 600 sites including Tribune Publishing and had signed BP as its first non-media customer, the proof point that the platform could sell beyond newsrooms.
Today's numbers show that bet compounding: 2,000+ licensed companies, up from about 1,500 a year ago, with a much larger share of clients outside media and entertainment. What hasn't compounded yet is margin — Axios reports roughly $40M–$50M in ARR against ongoing losses, which is what turned the unit's ownership into a live question within months of this report.
First-order effects
- The Post is carrying an unprofitable software unit inside a news company: ~$40M–$50M in recurring revenue across 2,000+ licenses does not yet cover the cost base of the ~250-person global operation the platform runs on.
- Client growth is skewing toward non-media brands, confirming the BP signing as the start of a diversification rather than a one-off — media remains the largest base but is no longer the growth engine.
Second-order effects
- Every publisher that licenses Arc converts its own infrastructure build cost into WaPo revenue — Tribune Publishing was already running on the platform by 2019 — pushing newsroom technology economics from build to buy.
- A larger non-media installed base makes the second half of the 2018 plan, an ad network riding Arc's footprint, more plausible, since reach would extend well past The Post's own audience.
Third-order effects
- If licensing keeps scaling while the unit stays loss-making, the logical endpoint is separating the tech arm from the newsroom — months later, sources told the Journal WaPo was exploring a spinoff or sale, with projections of $200M+ ARR by 2027.
- The structural shift is newspapers becoming software businesses: WaPo's path from internal CMS to a product sold to thousands of companies is the template for legacy media monetizing engineering rather than only content.
The trend: Legacy newsrooms are converting internal publishing infrastructure into standalone software businesses, with profitability pressure increasingly forcing decisions about whether those units stay owned.