Imagine the person who co-built a category coming back to replace it. That’s essentially what’s happening right now in marketing automation. Jon Miller, co-founder of Marketo, has launched a new company called Phave – explicitly positioning it as the replacement for the rules-based marketing automation systems he helped create two decades ago. At almost the same moment, Omneky has wired its agentic advertising platform directly into ChatGPT, and HubSpot has claimed a Leader position in the 2026 Gartner Magic Quadrant for B2B Marketing Automation Platforms. The category isn’t static. It’s mid-reinvention.

For RevOps and CRM teams who’ve spent years configuring drip sequences and lead scoring models, this week’s news carries a specific implication: the mental model you use to think about marketing automation may already be out of date.

What Phave Is Actually Challenging About Marketing Automation Platforms

Phave’s core argument is direct. Today’s marketing automation platforms – Marketo, HubSpot, Pardot, and their peers – are fundamentally built on if-then logic. A contact hits a trigger, a workflow fires, an email sends. It’s deterministic by design. Phave wants to replace that with campaigns that reason, adapting in real time based on intent signals and context rather than executing a pre-scripted branch.

That’s a genuinely significant architectural shift. Rules-based systems are predictable and auditable, but they’re also brittle – they do exactly what you told them to do, even when customer behavior moves in a direction the workflow designer didn’t anticipate. An AI-native system that reasons through decisions rather than following a flowchart could handle edge cases more gracefully. But it also introduces questions about explainability and control that CRM teams will need to get comfortable with fast.

Miller’s credibility here is hard to dismiss. He didn’t just use Marketo – he built it. His follow-on company, Engagio, tackled account-based marketing before being acquired by Demandbase. Phave is his third attempt to define what the category should look like at its next stage, and the framing this time is unambiguous: the platform he built is now the problem he’s solving.

Omneky Brings Agentic GTM Execution Into ChatGPT

While Phave takes aim at the campaign logic layer, Omneky is attacking the execution layer from a different angle. The company announced this week that its agentic harness for advertising and go-to-market programs is now accessible directly through ChatGPT. In practice, a marketer could run a brand’s entire multi-channel ad program through a single conversation interface, with the agent pulling from the brand’s own customer data to make decisions.

That’s worth pausing on – not because conversational interfaces are new, but because the data grounding is what changes the calculus. An agent operating against a brand’s actual customer data – purchase history, segment behavior, prior campaign performance – is doing something meaningfully different from a generic AI tool generating ad copy on request. It’s closer to having a strategist who has already read every report in your Ideal Customer Profile brief before the conversation starts.

The GTM implications extend beyond advertising. If agentic systems can execute across channels from a single interface grounded in proprietary data, the role of the campaign manager shifts – less configuration work, more judgment calls about what the agent should prioritize and why. As we covered in How Agentic AI Is Quietly Rewiring the GTM Stack, this pattern is showing up across the stack, and Omneky’s ChatGPT integration is one of the cleaner real-world examples of it becoming a product reality rather than a demo.

HubSpot’s Gartner Recognition and What It Signals

HubSpot’s placement as a Leader in the 2026 Gartner Magic Quadrant for B2B Marketing Automation Platforms matters for a specific reason: it arrives exactly when the category is being contested most aggressively. Gartner Leader status is a procurement signal – what shows up in evaluation shortlists and board decks. Holding that position while newer entrants like Phave argue the category needs to be replaced is a meaningful competitive advantage for HubSpot’s sales motion.

It also reflects something real about HubSpot’s trajectory. The company has been methodically expanding its AI capabilities while keeping its CRM and marketing automation tightly integrated. For teams that want a mature, well-supported marketing automation platform with AI features building on top of a stable foundation, that Gartner recognition will carry weight in the buying process. You can explore how it stacks up against alternatives in our Tool Reviews section.

Context Engineering: The Hidden Fight Underneath All of This

There’s a less visible conflict running beneath the Phave launch and the Omneky announcement, and it’s arguably more consequential for how CRM teams should think about their infrastructure. CMSWire’s coverage this week highlighted why context engineering is becoming the most contested control point in martech – and the logic is sharp.

Marketing AI systems are only as good as the context they receive before generating a response or taking an action. That context comes from somewhere: your CRM, your CDP, your data warehouse. Whoever controls the layer that assembles and delivers that context to the AI effectively controls what the AI does. It’s why CRM vendors, CDP providers, and warehouse platforms are all moving to own this layer simultaneously.

• CRM vendors argue they hold the most complete customer record and should be the system of context for any AI operating in the GTM stack.
• CDP platforms point to their real-time identity resolution and behavioral data as the richer, fresher signal set.
• Data warehouse vendors are positioning their platforms as the neutral, high-volume source of truth that AI should be trained against and queried from.

This isn’t a theoretical turf war. It has direct consequences for sales cycle length, Customer Acquisition Cost, and campaign efficiency. If your AI agent is working from stale or incomplete context, it’s making decisions based on a partial picture – and that shows up in performance metrics before it shows up in vendor conversations.

The Salesforce-SAP API Dispute Adds Another Layer of Complexity

Separately, The Register reported this week that experts are questioning whether a recent Salesforce demo breached SAP’s API policy. The specifics are still being evaluated, but the broader concern matters here: as major CRM and enterprise software vendors push to control more of the user experience layer – including what data they can surface from adjacent systems – API access and data portability become live operational issues, not hypothetical ones.

For teams running complex GTM stacks that span multiple enterprise platforms, this kind of dispute is a preview of conflicts that could affect how freely data flows between systems. It’s the same underlying tension showing up in the context engineering debate: who controls the data layer, and on what terms. If you’re building or evaluating a stack right now, this is worth tracking closely. Our guide on How to Build a GTM Stack That Survives Platform Power Grabs covers the strategic framing in detail.

What CRM and RevOps Teams Should Actually Do With This

This week’s developments don’t require immediate action. But they do require clear thinking about where your current stack sits relative to what’s coming. A few specific questions worth asking now:

• Is your marketing automation platform built on rules-based logic that your team actively maintains, or does it have genuine reasoning capabilities that adapt without manual intervention?
• Where does the context that feeds your AI tools come from – and do you actually control that layer, or is a vendor assembling it for you behind the scenes?
• If an agentic system were operating your ad spend across channels today, what data would it be grounded in – and how fresh is that data?
• How exposed is your stack to API policy changes between platforms you depend on for data flow?

None of these questions have tidy answers yet. Phave is early-stage. Omneky’s ChatGPT integration is brand new. The context engineering fight is ongoing. But the direction is clear enough to start auditing your assumptions now rather than waiting for a forcing event.

For teams managing sales pipeline and campaign attribution simultaneously, the convergence of these trends is worth particular attention. The systems that drive top-of-funnel demand are being rebuilt at the architecture level – not just updated with AI features bolted on. That’s a different kind of change, and it warrants a different kind of evaluation process. Check our CRM Guides for frameworks on how to assess your current setup against emerging alternatives.

If you want to stay ahead of how these shifts are playing out week by week, the CRM Daily Newsletter covers this beat closely.

The person who built the old playbook is now betting against it. That’s your signal.