When Automattic first announced plans last August to move all of Tumblr’s half‑billion blogs onto the WordPress infrastructure, the tech community buzzed with talk of seamless backend synergies and a potential gateway into the fediverse via WordPress.com’s ActivityPub plugin. Fast‑forward nearly a year, and Automattic CEO Matt Mullenweg has confirmed that the project is now “on hold” for an indefinite stretch, shifting the spotlight back onto features that users can actually see and feel.

On the latest episode of The Verge’s Decoder podcast, Mullenweg admitted that the migration was largely an infrastructure exercise—a big re‑architecture that, while “cleaner,” didn’t translate into immediate improvements for the end user.

“What we decided is that we want to focus as much on the things that are going to be noticeable to users and that users are asking for,” he said. “This was more like an infrastructure thing… I still want to do it. It’s just cleaner. But right now, we’re not working on it.”

That emphasis on visible, user‑driven features neatly aligns with Automattic’s broader restructuring earlier this spring, when the company announced layoffs amounting to some 16% of its workforce, including portions of the Tumblr team, to sharpen its focus on core products and customer‑facing innovation.

One of the migration’s biggest selling points was Tumblr’s potential entry into the fediverse—a decentralized social web where platforms like Mastodon, PixelFed, and WordPress.com can freely exchange posts via ActivityPub. The plan was simple: once Tumblr ran on WordPress infrastructure, posts could automatically federate without building new protocols from scratch.

But with the migration shelved, fediverse integration is put on ice too. Mullenweg suggested that if there’s grassroots demand for fanning out Tumblr into the fediverse, Automattic could add ActivityPub support directly into the existing Tumblr code base rather than waiting on WordPress. It’s a pragmatic pivot, but one that leaves the fediverse crowd wondering if Tumblr will ever truly decentralize.

Tumblr’s journey hasn’t been smooth sailing. Acquired by Yahoo in 2013 for $1.1 billion, then sold to Automattic in 2019 for under $3 million, the platform has weathered revenue challenges, policy overhauls, and shifts in user tastes. Under Automattic, Tumblr gained new features—like live video streams and tipping—and dabbled in selling user data to AI firms, moves that sparked both excitement and controversy.

Yet profitability remains elusive. Mullenweg reiterated on Decoder that Tumblr “is not profitable” and continues to run thanks to cross‑subsidies from other Automattic products. Still, he insisted the platform has a future and deserves runway to experiment:

“I believe in Tumblr’s vision. We’re going to give it the time it needs to figure things out.”

With the WordPress migration parked and fediverse integration on standby, Tumblr’s roadmap is once again wide open. In the short term, users can expect a trickle of incremental updates—UI tweaks, feature refinements, and community‑driven additions that don’t require massive architectural overhauls. Long term? That remains anyone’s guess.

For now, Automattic is doubling down on what its millions of users actually see and interact with, rather than the plumbing that runs beneath. Whether that translates into revived growth or simply a holding pattern depends on how well Tumblr can deliver sticky, engaging experiences—and whether Automattic ever circles back to the dream of a federated, WordPress‑powered Tumblr.