Rémi KokabiNovember 4, 20266 min read
AdKit integration with lemlist through MCP, with prompts to copy
Your lead opens the cold email you sent through lemlist, then closes it without replying. An hour later they scroll past three ads from competitors and never see yours, because the AdKit audience your ad team built has no idea that lead is mid-sequence. Two tools, same prospect, zero coordination.
That gap is what the adkit lemlist integration through MCP closes. Connect both MCP servers in the same AI client and an agent reads your lemlist campaigns, builds matching audiences in AdKit, and keeps ads and outbound pointed at the same people without a manual export.
In this article I will show you exactly how to set that connection up and give you prompts you can paste today to show ads to leads mid-sequence and retarget the ones who never replied. You will also get the guardrails to keep an agent from making changes you did not approve.
Nobody connects outbound and paid the way they should, mostly because moving data between the two takes too much manual work. MCP removes that excuse.
Why connect lemlist and AdKit through MCP
Connecting lemlist and AdKit through MCP lets an AI agent coordinate your outbound sequences and paid ads from a single conversation. Instead of exporting CSVs between tools, the agent reads your lemlist campaigns and builds matching audiences in AdKit so the same prospects see your ads while your sequence runs.
Prospects reply more when they already recognize your name. That is the core idea behind surround sound marketing: pair multichannel outreach with targeted ads so every touchpoint reinforces the last. When lemlist and AdKit share the same AI client, the agent can pull your sequence leads, push them into ad audiences, and keep both systems in sync without a single manual export.
An MCP server is a standardized layer that exposes a tool’s actions to an AI agent over a simple protocol. The agent discovers what each server can do, calls the right actions, and chains results across servers in one conversation. You can learn more about how this works for sales teams in this guide to MCP for sales.
What you need before you start
Three things should be in place before you paste your first prompt:
The lemlist MCP server connected in Claude, ChatGPT, or Cursor. Follow the steps in how to set up the lemlist MCP if you haven’t done this yet.
The AdKit MCP server connected in the same AI client. AdKit uses OAuth, so you sign in once and the connection is live.
Permissions scoped to what each use case needs. Only grant write access when you are ready to act on results.
Start with read-only prompts. Ask the agent to list campaign leads or pull audience data before you let it create anything in either tool. That way you confirm both servers respond correctly and the data looks right before anything changes.
1. Show ads to leads in a sequence
Most teams run ads and outbound as separate motions. The ad team targets one audience, the SDR team emails another, and the prospect never sees both. That gap means your outbound leads hit a cold inbox with zero brand recognition, even when you run account based prospecting plays that should feel coordinated.
Paste this prompt into your AI client with both MCP servers connected:
Build a LinkedIn and Meta audience from the leads in campaign X so they see our ads while the sequence runs.
The agent reads the lead list from your lemlist campaign, extracts the identifiers it needs (emails, company names), and creates matching custom audiences in AdKit for LinkedIn and Meta. From there, your ads run alongside the sequence so every lead sees your brand before and after each email or LinkedIn touch.
Before running this on a full campaign, test on a small batch and confirm the audience size in AdKit matches the lead count in lemlist.
2. Retarget leads who did not reply
A sequence ends, the lead never replied, and the contact sits in a finished campaign doing nothing. That is pipeline left on the table. Retargeting those leads with ads keeps your name visible while you decide whether to re-engage with a new sequence.
Paste this prompt:
Take leads who finished campaign X without replying and add them to a retargeting audience for the next 30 days.
The agent filters your lemlist campaign for leads with a “finished without reply” status, pulls their details, and adds them to a retargeting audience in AdKit. You set the ad creative and budget on the AdKit side. The audience stays active for 30 days, giving you a window to warm those leads back up.
Before running this on a full list, verify the agent is filtering the correct status. Ask it to show you the filtered leads first, then approve the audience push.
Guardrails when an agent works in lemlist and AdKit
Letting an agent write in two tools at once is powerful, but it needs boundaries. Keep these in place:
Review every campaign and audience before launch. AdKit uses a draft-first workflow, so nothing publishes without your approval. Apply the same discipline on the lemlist side.
Test on 10 leads first. Run every new prompt on a small batch, check the output, then scale.
Keep unsubscribes and exclusion lists active. The agent should never override opt-outs in lemlist or add excluded contacts to an ad audience.
Limit who on the team can connect the MCP servers. One or two people should own the connection and the permissions behind it.
Log what the agent changed. After each session, ask the agent to summarize every action it took in both tools. Save that log for your team.
Run it in lemlist
The lemlist MCP server works in Claude, ChatGPT, and Cursor. Once connected, the agent can do far more than sync audiences.
It can search the lemlist People database (650M+ contacts, 65M+ companies) to find prospects that match your ICP, then enrich them with waterfall enrichment across 14+ providers to get verified emails and phone numbers. From there, it can create campaigns and sequences across email, LinkedIn, calls, and WhatsApp, all from the same conversation.
The agent also reads your unified inbox and campaign stats, so it can flag underperformers, suggest copy changes, and act on intent signals like job changes, funding rounds, or hiring activity. When you pair that with AdKit audiences, you get a loop: find leads, show them ads, run the sequence, retarget the ones who don’t reply, and re-engage when a new signal fires.
You can explore more workflows in the full list of lemlist MCP use cases.
Final takeaway
Connecting AdKit and lemlist through MCP turns two separate tools into one coordinated motion: ads and outbound hitting the same people at the same time. The prompts above work today in any AI client that supports both MCP servers. Start with a single campaign, test on a small batch, and scale once the data checks out.
Launch AI outbound campaigns
Automate prospecting and personalize AI-powered sequences to book more meetings.