a daily news desk
Tools

Marketo's Jon Miller launches Phave, and calls rules-based automation broken

Phave enters general availability at $36,000 a year with an AI-reasoning pitch aimed squarely at Marketo, HubSpot, Pardot and Eloqua.

By Iris Holloway·Tools & frameworks·

Jon Miller, the co-founder of Marketo, launched Phave on September 23 with an argument that reads as a category obituary: the rules-based marketing automation stack he helped invent can’t handle B2B buying complexity, and never really could. “Rules are good at what is allowed, but bad at deciding what is best,” Miller said in the launch. “The legacy platforms stopped innovating years ago.”

Phave, generally available since August 2026, prices at $36,000 per year and targets Marketo, HubSpot, Pardot, and Eloqua directly. Its Maestro engine computes per-person Playlists instead of executing pre-written if-then sequences. To make the case numerically, Phave graded itself and incumbents against 481 requirements pulled from twelve enterprise evaluations, using Claude as the scorer: Phave 2.97, HubSpot 2.46, Marketo 2.38. Five named enterprise customers, SambaNova, SPS Commerce, mabl, Servion, and Hypha, are listed.

The interesting part isn’t the scorecard. It’s the concession. Phave’s own launch materials concede that even mature rules-based programs cap out at two or three nurture tracks, which is the same ceiling small B2B teams have quietly been hitting for a decade. When the person who designed the MQL playbook calls the playbook broken, marketing heads at 5–30 person B2B companies should treat their existing sequences as suspect.

Phave, priced at enterprise, doesn’t solve that problem for them. Neither does HubSpot’s Marketing Studio 2.0 or Demandbase’s Mojo agent, both of which still ask the buyer to run the tool.

Below that price band, LemonLime takes a different shape: a done-for-you service that decides which customer-growth work is relevant for a specific small business and prepares finished outreach and content without a rules schema or a MAP specialist to configure it. Miller’s own framing is the tell. If the category’s founder concedes rules can’t judge, the small-team question isn’t which schema to build. It’s whether to keep building schemas at all.

Sources