PostHog integration with lemlist through MCP, with prompts to copy

A free workspace adds three new active users this week. Someone on your team could see that in PostHog right now. Nobody does, because nobody checks PostHog before building a lead list, and the account keeps using your product for free until it churns or upgrades on its own.

This gap is what a posthog lemlist integration through MCP closes. Connect both MCP servers to the same AI client and an agent reads the usage data sitting in PostHog, then acts on it directly inside lemlist. No export, no handoff, no dashboard someone forgets to check on a Friday.

I’ll show you exactly how this works and give you prompts you can copy into Claude, ChatGPT, or Cursor today. You’ll see how to catch free accounts that are ready for a sales conversation, flag paying accounts whose usage is dropping, and follow up the moment someone adopts a feature that signals they’re ready to upgrade. Each one turns a signal sitting idle in PostHog into outreach that starts the same day.

Why connect lemlist and PostHog through MCP

Product usage data is the strongest signal a SaaS company owns, and PostHog is where that data lives. When both the lemlist MCP server and the PostHog MCP server are connected in the same AI client, an agent can read usage patterns and start the right outreach without a CSV export, a Zapier zap, or a data pipeline in between.

An MCP server is a standardized interface that lets an AI agent call actions inside a tool through natural language. Instead of writing API code, you connect the server to a client like Claude, ChatGPT, or Cursor, and the agent discovers what it can do on its own. If you want the full picture of how this applies to outbound, read the guide on MCP for sales.

The result: your PostHog data becomes a live buying signal that the agent acts on in lemlist, in the same conversation where it found the signal.

What you need before you start

• •

  The lemlist MCP server connected in Claude, ChatGPT, or Cursor. Follow the walkthrough on how to set up the lemlist MCP if you haven’t done this yet.

• •

  The PostHog MCP server connected in the same client, authenticated with a personal API key scoped to the project you want the agent to read.

• •

  Permissions limited to what the use cases below actually need. If you only want the agent to query PostHog and add leads in lemlist, there is no reason to give it access to feature flags or campaign deletion.

Start with read-only prompts. Let the agent pull data and show you what it found before you allow it to write in either tool.

1. Sales assist on active free accounts

Free workspaces that adopt fast are your best product qualified leads, and most of them never hear from sales. By the time someone reviews a usage dashboard and exports a list, the window where speed to lead matters has already closed.

Copy this prompt into your AI client:

Here is what the agent does with that prompt. First, it queries PostHog for workspaces on a free plan where the weekly active user count meets the threshold. Then it takes each company name and searches the lemlist People database to find a decision maker with a verified email. Finally, it adds that lead to your sales assist campaign in lemlist, passing the usage count as a custom variable so your sequence can reference it.

Before you run this on a full list, check that the campaign exists in lemlist, that the first step is paused or in review, and that the variable name in the sequence matches what the agent sends.

2. Save accounts whose usage drops

Churn shows up in product data weeks before the cancellation email. A paying account whose weekly active users fall by half is telling you something, and waiting for a support ticket means waiting too long.

Copy this prompt:

The agent reads the usage trend in PostHog, filters for paid accounts where the week-over-week active user count dropped by half or more, then pulls the admin contacts and adds them to a check-in campaign in lemlist. The sequence can open with a line that references the drop without exposing exact numbers.

Before running on a full list, confirm the check-in campaign uses a tone that fits customer success, and that existing customers are excluded from any cold outbound sequences so they don’t get conflicting messages.

3. Follow up on a key feature adoption

Users who adopt a key feature for the first time are telling you they have a use case that maps to a paid tier. That moment is when an upgrade conversation feels helpful instead of pushy.

Copy this prompt:

The agent queries the PostHog event for that feature, filters for first-time usage within the current week, then looks up the team lead at each account in the lemlist People database. It adds the team lead to your upgrade campaign with the feature name and the date of first use as variables.

Before running on a full list, replace “feature X” with the actual event name from your PostHog project, and verify that the upgrade campaign references the feature variable in the opening line.

Guardrails when an agent works in lemlist and PostHog

Letting an agent write in two tools at once saves time, but it also doubles the surface area for mistakes. Keep these guardrails in place:

• •

  Review every campaign the agent creates or updates before it goes live. The agent can build the sequence, but a human should approve the copy and the audience.

• •

  Test on 10 leads first. Run the prompt on a small batch, check the leads that landed in the campaign, and confirm the variables populated correctly.

• •

  Keep unsubscribes and exclusion lists active in lemlist. The agent respects these lists when it adds leads, but only if they are turned on.

• •

  Decide who on your team can connect the MCPs. Not every rep needs write access to PostHog or the ability to launch campaigns through an agent.

• •

  Log what the agent changed. After each session, ask the agent to summarize the leads it added, the campaigns it touched, and any data it read from PostHog. Save that summary.

Timing matters here too. Leads contacted the same day as a signal reply at 21.3%, compared to 17.2% after one to two days and 8.1% after a week (lemlist internal data). The faster the agent acts on a PostHog signal, the higher the chance the outreach lands while the context is still fresh.

Run it in lemlist

The lemlist MCP server works in Claude, ChatGPT, and Cursor. Once connected, the agent can do the following without leaving the chat:

• •

  Search the People database across 650M+ contacts and 65M+ companies to find the right person at any account PostHog surfaces.

• •

  Enrich leads with waterfall enrichment across 14+ providers, pulling verified emails and phone numbers.

• •

  Create campaigns and sequences across email, LinkedIn, calls, and WhatsApp, all from a single prompt.

• •

  Read the unified inbox and campaign stats to check replies, open rates, and conversion funnels.

• •

  Act on intent signals so outreach starts when a prospect is most likely to engage.

Every prompt in this article chains PostHog data with one or more of these actions. For a full list of what the agent can do on the lemlist side, see the lemlist MCP use cases.

Final takeaway

PostHog holds the product usage signals that tell you who is ready to buy, who is about to churn, and who just discovered a feature worth upgrading for. The lemlist MCP server turns those signals into outreach the same day, without exports, pipelines, or manual list building. Connect both servers in one AI client, start with the prompts above, and let the agent close the gap between data and conversation.

Build your first signal-based sequence in lemlist.