Contentsquare and Dust announced an integration on June 24, 2026, that connects Contentsquare's behavioral analytics platform directly to Dust's enterprise AI agent infrastructure through a Model Context Protocol connector, allowing teams to query live customer experience data through natural language without leaving their existing workflows.

What the integration does

The pairing brings together two distinct layers of the enterprise AI stack. Contentsquare, headquartered in Paris and active across more than 1.3 million websites globally, captures granular behavioral signals - step completion times, drop-off rates, JavaScript error rates, API error rates, and session-level frustration scores. Dust, founded by Gabriel Hubert, provides the agent platform that enterprise teams use to build, deploy, and coordinate AI workflows. The new connector sits between them, making Contentsquare's datasets available as a live data source inside any Dust agent.

According to Contentsquare, the integration removes a friction point that commonly slows down cross-functional analysis. Understanding where visitors abandon a checkout flow - or quantifying the financial downside of a specific error - has typically required accessing dedicated analytics dashboards, pulling data, and passing findings between specialist teams. The MCP connector changes that sequence. Teams can ask questions in natural language and receive answers grounded in Contentsquare's behavioral datasets, without switching platforms.

The connector is available immediately for joint customers of both platforms.

The five use cases defined at launch

Contentsquare described five specific use cases activated by the connector at launch.

Funnel and journey analysis briefs are the first. Dust agents can automatically extract step completion times, drop-off rates, and end-to-end funnel conversion rates. The intent is to surface where visitors are lost so that product and engineering teams can prioritise fixes without running separate analyses.

Quantified revenue impact is the second. According to Contentsquare, Dust draws on Contentsquare's Impact Analysis to rank friction points and errors by their exact financial downside. That ranking is designed to help sprint planning teams prioritise remediation work based on measurable cost rather than intuition.

Automated behavioral digests are the third use case. Teams can configure autonomous, scheduled workflows inside Dust to deliver weekly or daily plain-English performance summaries to stakeholders via email. The workflow runs without human initiation once configured.

Error spike triage and routing is the fourth. The connector provides access to real-time JavaScript and API error rates alongside session impact metrics. When anomalies occur, Dust agents can generate immediate technical briefs or route triage notes directly into connected engineering systems such as Slack or Jira. This use case is particularly relevant for engineering and platform reliability teams who currently rely on separate monitoring tools.

Multi-source business reasoning is the fifth. The connector is designed to work alongside other data sources already connected to Dust agents, enabling the platform to provide a unified view of business insights across a broader technology stack.

How the Model Context Protocol enables this

The technical mechanism underlying the integration is the Model Context Protocol, an open standard that Anthropic introduced in November 2024. MCP defines how AI models communicate with external tools and data sources through a standardised interface. Rather than requiring custom integration code for each pairing, vendors publish an MCP server that exposes their data and functionality to any compatible AI client.

Contentsquare's implementation follows this pattern: the MCP connector functions as a server that exposes its behavioral datasets - session data, error rates, funnel metrics, impact analysis outputs - to Dust agents through a consistent interface. Dust agents query the connector at runtime, retrieving live data rather than operating on static exports.

The MCP standard has spread rapidly across the advertising and marketing technology industry since its introduction. HubSpot launched a CRM connector for Claude on July 29, 2025, enabling natural language access to customer relationship management data. Amazon Ads launched its own MCP server in closed beta on November 13, 2025, supporting connections from Claude, ChatGPT, Amazon Q, and Amazon Bedrock. Meta opened its advertising infrastructure to Claude and ChatGPT via MCP connectors on April 29, 2026. Lifesight launched an MCP connector on June 2, 2026, giving Claude and ChatGPT direct access to marketing measurement data for clients managing more than $4 billion in combined spend.

The Contentsquare-Dust integration follows the same structural logic. The bet is that teams are already using AI agents as their primary work interface; the more productive path is to bring the data to those agents, rather than asking users to open separate analytics platforms.

Dust's platform and enterprise position

Dust describes itself as an enterprise platform for building and deploying secure AI agents. According to Dust, more than 3,000 organisations use the platform. The system allows teams to create agents that are customised to company-specific workflows and to connect those agents to a range of data sources and tools - Notion, Slack, GitHub, and external websites, among others.

A central design feature of Dust is model flexibility. Teams can choose from OpenAI, Anthropic, Gemini, Mistral, or other foundation models and switch models when newer versions are released. The platform holds certifications covering SOC 2, HIPAA, and GDPR. According to Dust, data hosting location and data selection are both configurable within its security parameters, which is a relevant detail for enterprise procurement processes in regulated industries.

The platform positions itself not primarily as a chatbot interface but as infrastructure for what it calls AI Operators - people inside organisations who build and manage agent workflows on behalf of their teams. That framing distinguishes Dust from general-purpose AI assistants and situates it in enterprise workflow automation.

Gabriel Hubert, co-founder and CEO of Dust, described the rationale for the integration in Contentsquare's announcement: "Contentsquare's MCP connector is a great example of what happens when rich data meets AI agents: insights that used to require specialist teams and multiple dashboards are now available to anyone, in seconds. We're excited to bring that experience intelligence directly into Dust workflows."

Contentsquare's data layer and the context problem

What makes behavioral data particularly useful in an agentic context is that it answers a different question from performance data. Revenue metrics and conversion rates tell teams what happened. Behavioral data - where users paused, where they encountered errors, which elements they ignored, how long they spent on each step - provides evidence of why outcomes occurred.

Jonathan Cherki, CEO and founder of Contentsquare, articulated this directly in the announcement: "AI agents have changed how quickly we can get answers; what's often missing is context. When agents can access customer behavior data, they can help teams understand not just what's happening, but why - making their responses, briefs and workflows far more impactful."

That distinction is operationally significant. An agent with access only to aggregate conversion data can identify a problem but not diagnose it. An agent with access to Contentsquare's behavioral layer - frustration signals, error session impacts, step-by-step completion timing - can generate a brief that describes both the problem and its likely cause, in terms a product manager or engineer can act on directly.

Contentsquare's scale gives this a practical dimension. More than 1.3 million websites use the platform globally. The Shopify partnership announced December 17, 2025 extended behavioral measurement tools to Shopify's commerce infrastructure, which spans merchants across more than 175 countries.

Where this sits in the agentic marketing stack

The Contentsquare-Dust integration reflects a broader structural shift that has been building for roughly eighteen months. The pattern is consistent: data platforms that previously required dedicated interfaces are publishing MCP connectors that bring their data into whatever AI environment a team already uses.

Agentic ad tech's push to take over the buying layer, documented by PPC Land in the week ending June 18, 2026, included DoubleVerify embedding MCP support to allow Claude to query media quality data, and LiveRamp opening its data platform to third-party agent builders. Microsoft expanded its Advertising MCP server to open pilot on June 17, 2026, allowing businesses to build custom AI workflows grounded in live campaign performance data inside Claude, ChatGPT, and M365 Copilot.

The customer experience intelligence category has not been as visibly represented in this wave as advertising technology, but the underlying logic is identical. Cross-functional teams - product, engineering, ecommerce, marketing - already use AI agents for analysis and reporting. The friction is not in asking questions; it is in getting those questions answered with data that reflects actual user behaviour rather than generic knowledge.

Contentsquare noted in the press release that it is among the first experience analytics solutions listed in Anthropic's Claude Connectors Directory, signalling that the MCP connector architecture extends beyond the Dust integration itself.

Technical specifics of the connector

The connector accesses several categories of Contentsquare data. Funnel metrics include step completion times and conversion rates across defined journey segments. Drop-off rates capture where and at what frequency users exit a defined flow. Impact Analysis outputs rank friction points by their calculated financial effect. JavaScript error rates and API error rates are surfaced in real time alongside session impact metrics that quantify how errors affect user behaviour.

The error routing use case is technically notable. Current monitoring workflows typically require separate alerting infrastructure and manual triage to connect an error spike to its user impact. The Contentsquare MCP connector enables Dust agents to perform that connection automatically - detecting the anomaly, assessing its session impact, generating a technical brief, and routing it to Jira or Slack - without human intervention at each step.

The multi-source reasoning use case extends the connector's value further. Dust's architecture allows agents to draw simultaneously from multiple connected data sources. A Dust agent connected to both Contentsquare and a customer data platform, for example, could combine behavioral signals with demographic or transactional data to produce a more complete diagnostic view than either source provides alone.

Why this matters for marketing teams

The marketing relevance here is direct. Identifying friction in checkout flows, understanding the impact of a recent release on user behaviour, and quantifying which errors are costing revenue are tasks that sit at the intersection of marketing, product, and engineering. They require data that has historically lived inside analytics platforms accessible primarily to specialist teams.

The practical consequence is cycle time. If a campaign manager wants to understand why a promotional landing page is converting below expectations, the typical path involves raising an analytics request, waiting for a specialist to pull the data, interpreting a report, and aligning on next steps across teams. A Dust agent with the Contentsquare connector connected can compress that path considerably - not by replacing the human judgment involved, but by eliminating the data retrieval steps.

PPC Land's coverage of MCP adoption across marketing technology has documented that the structural bet behind MCP connectors is friction reduction: meeting people in the tools they are already using rather than requiring them to open separate platforms. The Contentsquare-Dust integration is a direct example of that pattern applied to customer experience data.

Timeline

  • November 2024 - Anthropic introduces the Model Context Protocol as an open standard for connecting AI assistants to external data sources and tools.
  • May 1, 2025 - Anthropic launches Integrations, connecting Claude to remote MCP servers across web and desktop for the first time, with 10 initial services including Atlassian, Zapier, and Cloudflare. PPC Land
  • July 29, 2025 - HubSpot launches the first CRM connector for Anthropic's Claude, enabling natural language queries against CRM data. PPC Land
  • November 13, 2025 - Amazon Ads launches a closed beta for its MCP Server, supporting connections from Claude, ChatGPT, Amazon Q, and Amazon Bedrock. PPC Land
  • December 17, 2025 - Contentsquare announces a partnership with Shopify, integrating session replay, heatmaps, and frustration scoring into Shopify checkout flows across merchants in more than 175 countries. PPC Land
  • January 12, 2026 - Anthropic launches Cowork as a research preview for Claude Max subscribers, extending file automation and connector access to non-developers on macOS. PPC Land
  • April 29, 2026 - Meta opens its advertising infrastructure to Claude and ChatGPT in open beta via MCP connectors and a companion CLI, the first time Meta allows third-party AI systems direct access to live advertiser accounts. PPC Land
  • June 2, 2026 - Lifesight launches an MCP connector giving Claude and ChatGPT secured access to live marketing measurement data for customers managing more than $4 billion in spend. PPC Land
  • June 17, 2026 - Microsoft expands its Advertising MCP server to open pilot with read-only access, enabling live campaign data queries inside Claude, ChatGPT, and M365 Copilot. PPC Land
  • June 18, 2026 - DoubleVerify embeds MCP support for Claude into its Open Connectivity pillar; LiveRamp opens its data platform to third-party agent builders via MCP. PPC Land
  • June 24, 2026 - Contentsquare and Dust announce the Contentsquare MCP connector for Dust, available immediately to joint customers, enabling live behavioral data access inside Dust AI agents.

Summary

Who: Contentsquare, a global customer experience intelligence platform used by more than 1.3 million websites, and Dust, an enterprise AI agent platform used by more than 3,000 organisations.

What: The two companies announced an integration on June 24, 2026, built on the Model Context Protocol. The Contentsquare MCP connector gives Dust AI agents real-time access to Contentsquare's behavioral datasets, including funnel conversion rates, drop-off data, revenue impact analysis, and JavaScript and API error rates. Five specific use cases were defined at launch: funnel and journey analysis briefs, quantified revenue impact, automated behavioral digests, error spike triage and routing, and multi-source business reasoning.

When: The announcement was made on June 24, 2026, from Paris. The connector is available immediately for joint customers.

Where: The integration operates within Dust's enterprise AI agent platform. Contentsquare is headquartered in Paris. Documentation and setup information are available at contentsquare.com and dust.tt.

Why: Behavioral analytics data has historically been accessible only through dedicated dashboards, requiring specialist teams to retrieve and interpret findings. The MCP connector moves that data into the AI agent environments where cross-functional teams already work, compressing the time between identifying a performance issue and getting a diagnosis with enough context to act on.