Braze Pushes AI Into Campaign Decisions as Phave Targets Marketo's B2B Base

Campaign decisioning becomes the new battleground

Braze released AI Decisioning Studio Go and related agent capabilities in beta on September 29, targeting campaign orchestration rather than just content creation. The move positions Braze against Salesforce's Campaign Agent (announced September 14) and HubSpot's Marketing Studio beta in a race to automate which messages get sent, when, and to whom—not simply what those messages say.

The differentiation matters because enterprise marketing stacks already have generative copy assistants. The buyer question is now which vendor can automate the operational decisions that marketing operations teams currently handle manually: send-time optimization, channel arbitration, suppression logic, and audience decisioning across lifecycle campaigns. Braze says Decisioning Studio Go lets marketers configure and deploy AI agents without dedicated data-science or engineering support, though the company has not disclosed pricing, incremental costs, or quantified performance benchmarks.

Salesforce Marketing Cloud's Campaign Agent is scheduled for general availability in October. HubSpot's Marketing Studio beta is available now for Marketing Hub Professional and Enterprise customers. Braze's capabilities are in beta, with the first two expected to reach general availability in October. Buyers evaluating any of these platforms should separate beta functionality from production-ready capability and require contractual clarity on availability dates, service-level commitments, audit logs, human-approval controls, and whether AI features carry usage-based fees on top of base licensing.

Marketo co-founder launches Phave to replace linear lead nurture

Jon Miller, Marketo's co-founder, and Nick Bonfiglio, Marketo's former EVP of global product, brought Phave out of stealth on September 23. The platform targets Adobe Marketo Engage, Salesforce Account Engagement (Pardot), Oracle Eloqua, and HubSpot by replacing MQL-based nurture with buying-group behavior tracking. Phave reached general availability in August and lists SambaNova, SPS Commerce, mabl, Servion, and Hypha as customers. Its starting price is $36,000 per year, calculated by the number of people who receive something from the platform monthly.

The $36,000 floor sits above entry-level SMB automation tools but potentially below the total cost of a large Marketo or Eloqua deployment once implementation, data operations, and specialist administration are included. Miller's Marketo history provides category expertise, but the five publicly named customers and undisclosed funding amount—backed by FirstMark, Ridge Ventures, and Costanoa Ventures—place Phave in emerging-vendor territory. Buyers should request proof of integration with CRM, data warehouses, advertising platforms, consent systems, and account-level identity resolution. The buying-group orientation may require changes to existing MQL routing, attribution, and sales-acceptance processes.

Procurement teams should treat Phave as a competitive signal rather than a proven alternative. Enhanced diligence is required around financial continuity, migration tooling, contractual data portability, uptime history, and support capacity. The platform's differentiation is architectural—tracking buying groups instead of individual leads—but no independently verified retention, pipeline-contribution, or campaign-performance benchmarks are available.

Adobe acquires Rilo, shuts down standalone product

Adobe acquired Rilo, an AI workflow-automation startup, on September 2 and shut down Rilo's standalone product while integrating its capabilities into Adobe's marketing platform. Financial terms were not disclosed. Rilo had raised approximately $1 million from Peak XV, DeVC, and Day Zero Ventures at a reported $10 million valuation and reached more than 10,000 users within months of launch before the acquisition.

The transaction reinforces two buyer considerations. First, large marketing-cloud vendors including Adobe, Salesforce, Braze, and HubSpot are acquiring specialized AI workflow capabilities rather than building every agent function internally. Second, customers of small AI marketing tools face product-continuity risk. Rilo's standalone shutdown demonstrates that a product with more than 10,000 users can disappear as an independent offering after acquisition.

Adobe customers may gain deeper agentic workflow capabilities within Experience Cloud, potentially reducing the need for point tools. Buyers considering Rilo-like startups should require explicit contractual provisions for product deprecation notice, export formats, API continuity, and model and workflow portability. Enterprises should distinguish between vendors building AI features internally and those relying on acquired teams or external models, since acquisition-driven roadmaps carry integration and retention risk.

What to watch

Braze, Salesforce, and HubSpot are all scheduled to move AI decisioning capabilities from beta to general availability in October. Buyers should require vendor-run or customer-verified evidence on conversion lift, send-time optimization accuracy, suppression performance, and operational cost reduction before reallocating budget from existing campaign infrastructure. The competitive decision is shifting from which vendor has the best generative assistant to which one can reliably automate operational decisions that currently require human judgment.

Phave's pricing and buying-group architecture make it a credible challenger in B2B, but the small customer base and undisclosed funding limit its near-term suitability for risk-averse enterprises. Miller's Marketo pedigree matters, but procurement should prioritize evidence of scale, production reliability, and financial continuity over founder history. The platform's $36,000 starting price should be compared against the full cost of incumbent platforms, including implementation partners, data synchronization, and campaign operations—not just software licensing.