lemlist and HubSpot MCP use cases

A deal in HubSpot just got marked closed-lost. The contact still works at the company. Nobody in lemlist knows to reach out, because nobody checked. That gap between what your CRM knows and what your outreach tool does is where pipeline quietly dies.

This is what a proper hubspot lemlist integration is supposed to fix, and most teams still run it halfway. The native sync moves contacts and activities back and forth, but it can’t answer a question like “which champions changed jobs this month” or build a campaign from the answer. That takes an agent that can read both tools at once and act in the same breath.

In this article, I’ll show you exactly what the native integration handles versus what the MCP adds, and give you prompts you can copy into Claude, ChatGPT or Cursor today. You’ll see how to win back closed-lost deals, multi-thread stalled opportunities, and turn replies into pipeline, without exporting a single spreadsheet.

Native integration or MCP, what each one does

lemlist already syncs with HubSpot through a HubSpot integration that moves contacts, activities and statuses between the two tools in near real time. The MCP adds a different layer: it lets an AI agent query both HubSpot and lemlist in the same conversation and act on the answer right there.

The native sync handles the plumbing. Contacts created in one tool appear in the other. Emails sent, opened, clicked or replied to in lemlist log to the HubSpot contact timeline automatically. Field mapping controls which properties flow in which direction, and lifecycle changes in HubSpot can pause lemlist sequences so reps don’t cold-email a prospect already in a live deal.

The MCP does something the sync was never built to do. It lets you ask a question that spans both tools (“which closed-lost deals have contacts still at the same company?”) and get an answer, plus an action, in one prompt. The agent reads HubSpot through the HubSpot MCP server and acts in lemlist through the lemlist MCP server, all inside Claude, ChatGPT or Cursor.

Here is how the two compare side by side:

Keep both running. The native integration is your data backbone. The MCP is how you turn that data into outreach without exporting a CSV.

What you need before you start

Three things to have in place:

• •

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

• •

  The HubSpot MCP connected in the same client. HubSpot hosts a remote MCP server that works with Claude, ChatGPT and Cursor through OAuth.

• •

  Permissions scoped to what the use cases below need. Give the agent read access to deals, contacts and activities in HubSpot, and campaign and lead access in lemlist.

Start with read-only prompts. Ask the agent to pull data and show you the output before you let it create or update anything in either tool.

1. Win back closed-lost deals

Closed-lost deals sit in HubSpot with a loss reason attached and nobody assigned to bring them back. Most teams revisit them quarterly at best, if at all.

The agent reads the closed-lost deals and their loss reason from HubSpot, then checks each contact’s current company using lemlist’s enrichment. It creates a win-back campaign in lemlist, adds the leads, and maps the loss reason as a custom variable you can reference in the first email (“Last time, timing was the issue. Here’s what changed.”).

Before running this on a full list, review the campaign copy and confirm the loss reasons are clean enough to use as personalization. One malformed field can break the whole sequence.

2. Track champions who change jobs

A past buyer who moves to a new company is one of the warmest leads you can get. The problem is that nobody checks, so the signal passes without outreach. You can read more about why this matters in the guide on champion tracking.

The agent pulls every contact tied to a closed-won deal in HubSpot, creates a watch list in lemlist for job change signals, and links that watch list to a campaign so outreach fires when someone moves. Leads contacted the same day as a signal reply at 21.3%, against 8.1% after a week (lemlist internal data). Speed matters here.

Before running this on a full list, check that the campaign messaging references the prior relationship and the new role, not the old one.

3. Multi-thread open deals with one contact

Single-threaded deals stall when that one person goes quiet. You lose momentum, and nobody else at the account knows the conversation exists. The playbook for multi-threading in sales covers why this kills win rates.

The agent identifies single-contact deals above the threshold in HubSpot, searches lemlist’s People database for additional decision makers at the same company, creates those contacts in HubSpot, and adds them to a campaign that CCs the original champion. A first email to the person who runs the topic with the decision maker in CC replies at 3.35%, 23% more than a single recipient (lemlist internal data).

Before running this on a full list, verify the agent matched the right company and that the new contacts hold titles relevant to the deal.

4. Turn positive replies into deals

Interested replies get lost between the sequence and the pipeline. A prospect says “let’s talk” in lemlist, and nobody creates a deal in HubSpot until the rep remembers, sometimes days later.

The agent reads the lemlist unified inbox, filters for replies tagged as interested, creates a deal in HubSpot for each one, and attaches the reply text as a note on the deal record. The rep opens HubSpot and sees the deal with full context, ready to schedule a call.

Before running this on a full list, confirm the deal stage, pipeline and owner mapping so new deals land where your team expects them.

5. See which campaigns create pipeline

Campaign stats live in lemlist. Pipeline lives in HubSpot. Without connecting the two, nobody knows which outbound campaign actually generated revenue.

The agent reads campaign reply stats from lemlist and deal creation data from HubSpot, then matches them by contact email. The output is a ranked list of campaigns by total pipeline value, so you can double down on what works and cut what doesn’t.

Before running this on a full list, check that contact emails match across both tools. Mismatches from aliases or personal addresses will undercount attribution.

Guardrails when an agent works in lemlist and HubSpot

Letting an agent write in two tools at once is powerful, but it needs boundaries. Follow these before you scale any of the prompts above:

• •

  Review every campaign the agent creates before you launch it. The agent builds the structure, you approve the copy and the audience.

• •

  Test on 10 leads first. Run the prompt on a small batch, check the output in both tools, then expand.

• •

  Keep unsubscribes and exclusion lists active in lemlist. The agent respects them, but only if they’re turned on.

• •

  Decide who on the team can connect the MCP servers. Limit access to people who understand what the agent can change.

• •

  Log what the agent changed. After each session, check the campaign, the leads added, and any deals or notes created in HubSpot. If something looks off, you catch it before it reaches a prospect.

Run it in lemlist

Every prompt in this article runs through the lemlist MCP server, which works in Claude, ChatGPT and Cursor. Once connected, the agent can:

• •

  Search the People database (650M+ contacts, 65M+ companies) to find decision makers by role, industry, company size or tech stack

• •

  Enrich contacts with waterfall enrichment across 14+ providers for verified emails and phone numbers

• •

  Create campaigns and sequences across email, LinkedIn, calls and WhatsApp

• •

  Read the unified inbox and campaign stats to track replies, opens and conversions

• •

  Act on intent signals like job changes, funding rounds and hiring activity

These prompts are a starting point. You can explore more workflows in the full list of lemlist MCP use cases.

Final takeaway

The native HubSpot integration keeps your data in sync. The MCP lets an agent act on that data in one conversation, turning CRM records into campaigns, deals and attribution without a single export. Copy the prompts above, start with a small batch, and scale the ones that work.

Build your first signal-based sequence in lemlist.