Channel99, a San Francisco-based B2B marketing performance platform, today announced the integration of its Marketing Intelligence Data with three major generative AI systems - OpenAI's ChatGPT, Microsoft Copilot, and Claude Cowork - through a Model Context Protocol (MCP) server. The move makes Channel99 one of the first B2B marketing platforms to use MCP as a conduit for live cross-channel performance data flowing directly into commercial AI tools.
The integration, which became available immediately on February 24, 2026, allows marketing leaders to query their own pipeline data through conversational prompts, rather than navigating separate dashboards or exporting reports into spreadsheets. According to the company's announcement, the capability compresses the martech stack and opens what Channel99 describes as a path toward greater automation - though the practical scope of that automation will depend heavily on how organisations configure and adopt the new tools.
What the MCP server actually does
At a technical level, the integration works by running a Model Context Protocol server that sits between Channel99's data layer and the private instances of supported AI platforms. MCP, originally developed by Anthropic and donated to the Linux Foundation, functions as a standardised connection framework enabling AI applications to communicate with external data sources. PPC Land has tracked MCP adoption across the marketing technology industry since mid-2025, when Google began exploring the protocol for its Ads API. Since then, MCP has moved from experimental to mainstream, with companies such as AppsFlyer launching MCP tools in July 2025 and Google releasing an open-source MCP server for its Ads API in October 2025.
Channel99's implementation differs from these marketing analytics MCP releases in one important respect: rather than providing read-only reporting access to a single platform's data, the integration is designed to surface unified cross-channel B2B performance data - combining paid media, organic social, email, display, content syndication, intent signals, website engagement, and CRM systems - all at the account level. The company claims this architecture captures 10 times more customer signals than traditional attribution tools, specifically because it includes click-less engagement types that most attribution platforms have historically excluded.
The Marketing Intelligence Data Hub at the core of the product consolidates performance data across paid and organic B2B channels, connecting media platforms, intent data, website engagement, and CRM systems at the account level. Rather than relying on isolated platform dashboards, users interact with this unified data layer through simple conversational prompts in their preferred AI tool. The practical result, according to Channel99, is that a marketer can ask ChatGPT to recommend a budget allocation for a specific pipeline target, and the AI will draw on actual historical performance data - not generic guidance or training data alone.
Three functional areas are highlighted in the company's release. The first is LLM discoverability improvement: the system can identify top-performing keywords and topic clusters driving pipeline, with the aim of improving brand visibility within AI models and search engines. The second is intent-driven audience creation: users can prompt generative AI to build dynamic account lists based on historical performance data and receive recommendations for the optimal channel mix and budget. The third is outcome-based marketing plan generation: a marketer can specify a pipeline target and the system generates a complete marketing plan, including vendor selection and projected ROI.
"Customers want to engage with their data and performance insights through the tools they use every day, which are increasingly generative AI solutions," said Chris Golec, Founder and CEO of Channel99. "We are removing the guesswork and operational inefficiency, providing fact-based answers in seconds."
Golec previously founded Demandbase, an account-based marketing platform. Channel99's current product focus extends that background into cross-channel performance measurement, with a particular emphasis on connecting signals across the full B2B buying journey rather than single-platform attribution.
The attribution gap Channel99 is addressing
The context for this launch runs deep in B2B marketing measurement. Traditional attribution tools have largely been built around click-based signals - tracking which ad a person clicked before filling in a form or making a purchase. That model fits consumer e-commerce reasonably well. In B2B, it falls apart. Buying committees typically involve 6 to 10 people, sales cycles stretch across months, and many of the most significant brand interactions never generate a click at all. Display impressions, organic social posts viewed without engagement, email newsletters read but not actioned - these touch points influence decisions but leave no trail in conventional attribution systems.
PPC Land has reported extensively on how platforms have tried to close this gap. LinkedIn launched its Company Intelligence API in September 2025, enabling attribution partners including Channel99 itself to access aggregated company-level engagement data from LinkedIn's paid and organic touchpoints. Before that, LinkedIn had introduced company-level measurement in its Revenue Attribution Report in July 2025. B2B attribution has also been scrutinised in the broader context of last-touch attribution models making a contested comeback in some Google product contexts, a pattern that highlights the ongoing tension between measurement simplicity and accuracy.
Channel99's position is that none of these individual platform improvements address the underlying fragmentation problem - that data sits in disconnected systems and can only be meaningfully interpreted by analysts who know where to look and what to combine. By consolidating signals from across channels into a single account-level data layer, and then exposing that layer to AI through a standardised protocol, the company is making a structural bet: that the value of B2B marketing analytics lies not in any individual platform's data, but in the synthesis of all of it.
The claim of capturing 10 times the signal of traditional tools rests specifically on the inclusion of click-less engagement. Organic social engagement, email interactions, display impressions - these are captured at the account level, meaning that even when individual user identities are not tracked, activity from a given company is aggregated and associated with that organisation's pipeline status. This is the same theoretical approach that underpins LinkedIn's Company Intelligence API work, though Channel99's implementation is cross-platform rather than limited to a single network.
Security dimensions
The phrase "private instances" appears prominently in Channel99's announcement. The company specifies that the integration operates within private instances of ChatGPT, Microsoft Copilot, and Claude - meaning data does not flow through consumer-facing versions of those tools, but through enterprise deployments where organisations control their data environment.
This matters because MCP security vulnerabilities were identified by researchers in July 2025, with security analyst Akshay Pachaar characterising MCP security as "completely broken" in relation to tool poisoning attacks that exploit communication layers between MCP clients and servers. Channel99 does not address those vulnerabilities in detail in its announcement, referring instead to "secure, real-time access" as a feature of the product. The use of enterprise rather than consumer AI instances reduces - though does not eliminate - exposure to certain classes of risk identified in that research.
Competitive dynamics
The Ad Context Protocol, which launched in October 2025 with 23 participating organisations, represents one attempt to create standardised agentic AI workflows in advertising more broadly. Gracenote launched its own MCP Server in September 2025 for content discovery in television. Google launched a Universal Commerce Protocol in January 2026 for AI agent shopping workflows. Channel99's announcement lands in a landscape where MCP-based integrations are proliferating rapidly across marketing technology, with each company attempting to establish its data layer as the one AI tools should reach for.
What distinguishes Channel99's approach - and what makes it directly relevant to B2B marketing operations specifically - is the focus on pipeline rather than traffic or engagement. The core value proposition is not simply that AI can now access marketing data, but that AI can access marketing data that has already been tied to revenue outcomes at the account level. The quality of that tie, and how reliably Channel99's signals actually predict or explain pipeline, is the question B2B marketing leaders will need to evaluate when deciding whether this integration changes their operational workflow or simply adds a new layer to an already complex stack.
The integration is available immediately to existing Channel99 customers. New customers can access the functionality by creating a free account at channel99.com. The company is headquartered in San Francisco.
Timeline
- September 23, 2025 - LinkedIn launches Company Intelligence API for B2B attribution tracking, with Channel99 among five certified attribution partners at launch
- October 7, 2025 - Google releases open-source MCP server for Ads API integration, marking MCP's transition from exploration to implementation in advertising
- October 15, 2025 - Ad Context Protocol launches for advertising automation with 23 participating organisations
- November 2, 2025 - Ad Context Protocol coverage on PPC Land documents industry division on agentic AI standards
- December 11, 2025 - Channel99 launches one-click audience activation across LinkedIn, Google, Microsoft, Facebook, and YouTube
- January 11, 2026 - Google launches Universal Commerce Protocol with major retailers for AI agent shopping workflows
- February 24, 2026 - Channel99 announces the launch of its Marketing Intelligence Integration via MCP server, enabling access to its cross-channel performance data within private instances of ChatGPT, Microsoft Copilot, and Claude Cowork
Summary
Who: Channel99, a San Francisco-based B2B marketing performance platform founded by Chris Golec (formerly Founder and CEO of Demandbase), announced the integration. The integration supports users of OpenAI's ChatGPT, Microsoft Copilot, and Anthropic's Claude Cowork across enterprise deployments.
What: The company launched a Model Context Protocol server that provides real-time access to Channel99's unified cross-channel B2B marketing performance data inside private instances of three major generative AI platforms. The system captures click-less engagement signals across organic social, email, display, and content syndication - claiming 10 times the signal volume of traditional attribution tools - and ties that data to pipeline outcomes at the account level.
When: The announcement was made today, February 24, 2026, with immediate availability for current Channel99 customers.
Where: Channel99 is headquartered in San Francisco, California. The integration operates within enterprise AI environments - private instances of ChatGPT, Microsoft Copilot, and Claude Cowork - available globally to Channel99 customers.
Why: B2B marketing attribution has historically failed to capture click-less interactions such as display impressions, organic social views, and email reads, which heavily influence buying groups but leave no trace in conventional tools. By connecting a unified, pipeline-tied data layer to AI platforms through MCP, Channel99 is positioning itself as infrastructure for AI-generated marketing decisions - aiming to reduce the analytical burden on marketing operations teams and enable faster strategy execution through natural language interfaces.