Introducing Sellforte MCP: Bring Sellforte Into Your AI Workflows

AI is quickly becoming a new interface for how marketers work.

Today, working with Sellforte typically starts by opening Sellforte.

You explore your MMM results, review experiments, build a scenario, find the insight you need, and then take that information into whatever happens next.

But increasingly, marketing work is also happening inside LLMs.

Marketers are using AI to work across different data sources, analyze information, create presentations and documents, keep track of context, and connect different tools into the same workflow.

Today, we're introducing Sellforte MCP, our first step toward bringing Sellforte into those workflows.

Sellforte MCP makes Sellforte's marketing measurement and optimization platform accessible to LLMs and AI agents. That means your AI can use Sellforte alongside other information available to it, analyze marketing performance, create new optimizer scenarios when budgets change, and turn the results into whatever your workflow needs next.

The bigger idea is simple: you can now interact with your Sellforte data where you work, not only through the Sellforte UI.

Bring Sellforte into your AI workflow

Marketing decisions rarely depend on a single source of information.

Your MMM and experiment results might live in Sellforte. Campaign and other marketing data might be available through a source such as Supermetrics. Business targets could be in a document. Previous decisions might exist in conversations with your team.

An LLM can increasingly provide a common interface across those different sources.

Sellforte MCP makes Sellforte one of them.

MCP, or Model Context Protocol, provides a standardized way for AI applications to connect to external tools and data sources. With Sellforte MCP, an LLM can access the Sellforte tools and capabilities made available through the connector.

For example, if your AI environment has access to both Sellforte and marketing data provided through Supermetrics, you can work with them inside the same workflow.

Instead of manually exporting data from one system, combining it somewhere else, and moving the result into another tool for analysis, the LLM can work across the sources available to it.

The important part is not simply giving AI access to more marketing data.

It is giving AI access to Sellforte's independent measurement intelligence.

Give your AI access to MMM and Experiment data

Advertising platforms already provide enormous amounts of marketing data. Connecting that data to an LLM can make it easier to query and analyze.

But platform data alone does not tell you what actually drove incremental business impact.

Sellforte MCP makes Sellforte's MMM and experiment data available inside AI workflows, giving the LLM access to the independent measurement marketers already use in Sellforte.

Ask Claude to analyze your past media performance, for example, and it can use Sellforte to retrieve the relevant information and present the results directly in the conversation.

Ask a question in natural language and Claude can use Sellforte MCP to retrieve and analyze past media performance.

In this example, Claude compares media performance over the last 12 months with the prior year, using Sellforte to retrieve spend and incremental sales results across media groups.

From there, the analysis can continue.

Claude can use the same Sellforte data to examine how blended media ROI has developed over time, create a visualization, and explain what changed.

Sellforte data can be analyzed and visualized directly inside the AI workflow.

Here, Claude identifies that media spend decreased while incremental sales increased, resulting in an improvement in blended MMM-based ROI.

Claude creates the visualization and explanation and Sellforte provides the underlying marketing measurement.

That distinction is important.

The value of Sellforte MCP is not another generic AI analysis layer. It is giving the AI access to Sellforte's MMM, experiments, and measurement capabilities as part of a wider workflow.

--> Want early access? Join the Sellforte MCP waitlist.

Go from analysis to a finished presentation

Once Sellforte is available inside a general-purpose AI environment, the workflow does not have to stop when you find the answer.

Imagine you are preparing for a quarterly marketing review.

You can ask your LLM to analyze your Sellforte results.

Then ask it to explain the most important findings.

Ask which recommendations leadership should know about.

And finally, ask it to turn the analysis into a presentation.

After working with Sellforte data and recommendations, the same AI workflow can turn the results into a finished presentation.

In this example, Claude uses Sellforte media allocation information to create a six-slide executive presentation containing recommendations and supporting visuals.

The presentation is created using Claude's capabilities. Sellforte MCP provides the Sellforte information and planning outputs Claude uses as the foundation.

The same principle can apply to documents, PDFs, and other outputs supported by the AI environment.

This opens up use cases that do not need to exist inside the Sellforte product itself.

Sellforte provides the marketing intelligence. The LLM can take that intelligence wherever the workflow needs to go next.

Bring business context into the same workflow

There is another difference between working with Sellforte through an LLM and working inside a traditional analytics interface: the wider context available to the AI.

Marketing decisions are rarely made from numbers alone.

Perhaps your team decided that a particular market cannot absorb more budget. Maybe a campaign is intentionally being constrained because of an upcoming product change. Or leadership has already agreed on a business constraint that needs to guide the next planning cycle.

An AI environment can keep that context alongside the Sellforte analysis.

That can include other connected data sources, documents, conversations, and, where supported by the LLM, memory of previous decisions.

Instead of every analysis starting from zero, Sellforte's measurement can become one part of the broader context behind the decision.

This becomes particularly powerful when the business question changes.

From MMM analysis to MMM agents

Imagine leadership tells you that next quarter's media budget needs to be reduced by 10%.

Normally, that creates a new planning exercise.

Which channels should absorb the reduction? Which investments should be protected? And how much incremental sales impact should you expect from the cut?

With Sellforte MCP, you can ask the question directly:

"We need to cut media spend by 10%. Can you run a scenario and recommend how we best allocate our budget?"

Claude can use Sellforte's optimizer capabilities to create a new scenario based on that constraint and return the recommended allocation.

Claude uses Sellforte's optimizer to create a new media allocation scenario based on a 10% budget reduction.

The recommendation does not simply reduce every channel by 10%.

Sellforte's optimizer evaluates how the reduced budget should be allocated based on the underlying response curves. The resulting scenario shows where investment should change and the expected effect on incremental sales and ROI.

Claude can then interpret the results, explaining which investments should absorb larger cuts, which should be protected, and where investment could even increase despite the overall budget reduction.

The same workflow works in the other direction:

"We've received another €500,000 for Q4. Where should we invest it?"

Instead of manually translating each new business question into a separate planning exercise, the LLM can call on Sellforte when the question arises.

This is the beginning of what we think of as MMM agents: AI agents that can use Sellforte's measurement and planning capabilities to help marketers answer business questions and evaluate scenarios on demand.

From recommendation to action and back again

Getting to a recommendation is not the end of the workflow.

Today, Sellforte MCP can help bring measurement, analysis, and planning into LLM workflows. Supported recommendations can already be executed separately through Sellforte Activate across advertising platforms.

Coming soon, we want to connect these steps more closely.

An agent could identify a change in the business, use Sellforte to evaluate a new scenario, explain the recommended allocation, and then help move that recommendation toward execution through Sellforte Activate or connected advertising platforms.

The loop could then continue after the change is made.

Sellforte Change Intelligence can help marketers understand what happened next, evaluating the impact of campaign and bidding changes and adding that learning to the next decision.

The direction is a more connected agentic workflow:

Measure → analyze → plan → act → measure again.

This full workflow is not available through Sellforte MCP today, but it represents where we are heading: connecting independent measurement, planning, execution, and learning into a continuous decision loop.

How Sellforte MCP works

Our first implementation of Sellforte MCP uses Claude.

Inside Claude, Sellforte can be added as a Custom Connector.

Sellforte MCP can be added to Claude as a Custom Connector.

Connect Sellforte and authenticate using your Sellforte account.

Authenticate to Sellforte MCP using your Sellforte account.

Once connected, Claude can see the Sellforte tools and capabilities available through MCP.

The connector lists Sellforte capabilities including dataset discovery, data retrieval, optimizer scenarios, tables, and plots.

The current tools include capabilities for discovering and retrieving Sellforte data, creating and accessing optimizer scenarios, and rendering tables and visualizations.

You do not need to know which individual tool to call.

One of the easiest ways to start is simply to ask:

"What can you do with Sellforte?"

Claude can inspect the available Sellforte capabilities and help you explore them through natural language.

More technical users can also connect to Sellforte MCP through environments such as Claude Code.

Claude is our first working example but Sellforte MCP can be added to any LLM client that supports connecting to MCP servers.

Please note that adding the Sellforte MCP server requires a Sellforte subscription with the MMM module, plus an LLM client where you can configure an MCP server yourself or where your organization can add a custom connector for the OAuth-authenticated endpoint we provide.

The next step toward Agentic MMM

Sellforte MCP does not replace Sellforte.

It creates another way to interact with it.

There will continue to be workflows where the Sellforte product is the best place to explore results, understand response curves, work with experiments, validate findings, and build plans.

But there are also workflows where the question starts somewhere else, for example:

• A CFO changes the budget

• An analyst needs to combine Sellforte with another dataset

• A leadership presentation needs to be ready tomorrow

An agent needs independent measurement before it can make a useful recommendation and Sellforte MCP allows Sellforte's measurement and planning capabilities to become part of those workflows.

That is the bigger opportunity behind this launch.

Instead of Marketing Mix Modeling being something marketers always have to go somewhere to use, Sellforte's measurement can increasingly come to the question.

Join the Sellforte MCP Waitlist

Sellforte MCP is in early release and we're starting with a limited group of users.

We'll roll out access account by account so we can learn how customers use Sellforte inside their AI workflows, understand which use cases provide the most value, and continue improving the capabilities available through the connector.

If you're already a Sellforte customer and want to bring Sellforte into your LLM and agent workflows, join the waitlist and we'll be in touch as access expands.

👉 Join the Sellforte MCP waitlist

The Future of MMM Is Agentic

AI agents will only make good marketing decisions if they have access to good marketing measurement.

That makes independent measurement more important, not less.

Sellforte MCP is our first step toward making Sellforte's MMM, experiment data, and planning capabilities available wherever those decisions increasingly happen.