Will WebMCP become a ranking signal?

We explore whether Google is likely to incentivize agent-readiness across its surfaces, and what that would mean for marketers.

April 2026

Google's position is shifting

The share of information discovery that flows through Google is shrinking. SparkToro/Datos clickstream data showed Google desktop searches per user fell nearly 20% year-over-year between 2024 and 2025 (Search Engine Land, Jan 2026). Apple SVP Eddy Cue, testifying under oath in United States v. Google LLC in May 2025, said Safari search volume had declined for the first time in 22 years (Fortune). Buyers research in ChatGPT, Claude, and Perplexity. Agents browse the web without touching Google Search. The dynamics behind that shift are covered in why your organic traffic is down and the attribution problem nobody has solved.

Google is responding. In the past twelve months, Google has shipped AI Overviews, AI Mode, Project Mariner, the A2A protocol, UCP, a dedicated "Google-Agent" user-agent string, and WebMCP. Nick Fox, who leads Google's search product, recently stated that Google is "increasingly thinking of AI Mode and AI Overviews as one and the same" (Search Engine Journal, Mar 2026). Liz Reid, Head of Search, described a future where "a lot of agents are talking with each other."

All of these investments share a common structure: they keep agents interacting with the web through infrastructure Google controls. WebMCP, covered in detail in what is WebMCP, is one of the most significant, because it defines how agents interact with websites at the browser level, and Chrome is currently the only browser that has implemented it.

The HTTPS precedent

When Google wanted the web to move to HTTPS, it did not rely on a single incentive. It made HTTPS a search ranking signal in 2014. It added "Not Secure" warnings in Chrome in 2017. It reported HTTPS status in Search Console. It flagged insecure pages in Lighthouse, Google's site audit tool that scores pages on performance, accessibility, and best practices. The pressure came from multiple directions at once. By the time the transition was complete, HTTPS was effectively mandatory for any site that wanted to be taken seriously by Google's ecosystem.

Mobile-friendly followed the same pattern. Core Web Vitals followed the same pattern. Chrome shipped the measurement tools. Search provided the adoption incentive. Lighthouse made the score visible to every developer.

The question for WebMCP

Whether Google makes WebMCP a formal search ranking signal is one version of the question, and probably not the most important one.

Google Search and Google's AI surfaces are blending together. A user typing a query into Google increasingly interacts with AI Overviews or AI Mode rather than a list of blue links. In that context, the question is not narrowly about search ranking. It is about whether agent-readiness affects how a site shows up across any of Google's surfaces.

There is a range of things that could happen. WebMCP could become a formal search ranking signal the way HTTPS did. It could surface WebMCP-ready sites higher in AI Overviews without changing traditional rankings at all. It could become a preference signal in AI Mode, where agents route users toward sites they can interact with more reliably. It could show up as a Lighthouse metric that measures agent-readiness the way Lighthouse currently measures performance. Google could do several of these at once, the way it did with HTTPS.

Or agent-readiness could matter without Google ever formalizing it as a signal at all. If an agent in AI Mode is choosing between two competing sites to complete a task, and one site exposes structured tools through WebMCP while the other requires the agent to screenshot and guess its way through the interface, the agent has a practical reason to prefer the first site. That preference is a consequence of how agents work, not a policy decision.

What is observable

Google needs the agentic web to work. Its core business depends on the open web remaining the place where information lives and where transactions happen. If agents route around websites entirely, pulling information from APIs, caches, and closed platforms, that is bad for Google's ecosystem.

WebMCP, if adopted, keeps agent-site interaction inside the browser, inside Chrome specifically, where Google defines the API surface and captures the telemetry. Google has already invested in building the spec (co-authoring it through the W3C with Microsoft), shipping an experimental implementation, publishing demos, and opening an early preview program. Lighthouse already measures a range of web quality signals. Search Console already reports on structured data. The infrastructure for measuring and incentivizing agent-readiness largely exists.

Whether Google activates those levers for WebMCP specifically, and when, is what remains to be seen. WebMCP is still a Draft Community Group Report. Chrome has an experimental flag. No other browser has shipped an implementation. The spec could change. But the direction of investment is consistent, and the incentives are aligned: Google needs sites to be agent-ready, and Google has demonstrated, repeatedly, that it is willing to use its surfaces to drive adoption of standards it considers important.

Why this matters for marketers

Marketers who have been doing SEO for years already have a handle on search ranking. The more pressing question is what happens in the AI-mediated surfaces where discovery is actually shifting.

Whether agent-readiness becomes a formal signal, an informal preference, or just the practical reality that agents choose sites they can work with, the implication is the same: the sites that are easiest for agents to interact with will get more agent-driven traffic, and the sites that let agents take action will capture leads that invisible sites never see.

If you're interested in learning more, talk to us.