// CASE STUDIES

20 Meetings in One Month from 2 LinkedIn Profiles | AtomSell

Two LinkedIn profiles. One month. 20 meetings booked before a major medtech trade fair in Germany. Every flow behind it was built on AtomSell.

The client is a B2B company that sells to medtech manufacturers, distributors and service providers. Their Chief Commercial Officer was going to the fair in person and wanted her calendar full before she arrived. Their CEO was not travelling, but the CEO's profile ran in the same flows with a different angle. We built two AI agent flows on AtomSell and ran them from these two LinkedIn profiles.

The results

WhatResult
Meetings booked20
LinkedIn profiles used2
Period1 month
Connection requests sent1,286
Qualified leads23
Qualified leads that became meetings20 of 23 (87%)

Flow 1: the exhibitor list

The agent read the full exhibitor catalog of the fair, 4,066 companies, and kept 916 that matched seven ICP tracks: regulatory services, contract manufacturing, distributors, equipment service, manufacturers with dealer networks, startups, and education and training. For each company it found the LinkedIn page and the decision-makers: 1,133 contacts in 638 companies. Every message was written for the contact's track and for the sender, with a different voice for each of the two profiles.

Result: 8 meetings from 1,036 connection requests.

Flow 2: people who post about the fair

The second agent runs every day. It finds people who write on LinkedIn that they are going to the fair, checks them against the ICP and writes a message built from their own post. After the first weeks it was scaled:

  • search in 7 languages: English, German, French, Italian, Spanish, Dutch and Polish
  • a second, related fair added to the search
  • people who comment under posts about the fair
  • companies that announce their participation
  • decision-makers at the company when the post author has the wrong role

Daily output grew from 4 or 5 qualified leads to 35. These leads were sent first.

Result: 12 meetings from 250 connection requests.

Signals book 6.2X more meetings per request

FlowRequestsMeetingsMeetings per 100 requests
Exhibitor list1,03680.8
People posting about the fair250124.8

The signal flow used 19% of all connection requests and booked more than half of all meetings. The best combination was the client's Chief Commercial Officer, who was personally going to the fair, writing to people who had posted about it: 7 meetings from 134 requests, and every one of her 7 qualified leads booked a meeting.

Without AtomSell

The live-signal flow does not exist in today's stack. No tool finds who is posting about the fair in seven languages, decides who at that company to write to, and moves that lead to the front of a running campaign.

The exhibitor list is possible, but expensive: every one of the 4,066 rows classified and enriched with credits, and seven ICP tracks across two senders become 14 fixed campaigns to set up and maintain. On AtomSell it was one flow where the company, the person, the sender and the message were chosen for every lead, and any of them could change between batches without rebuilding anything.

What made it work

  • Two sources running in parallel: a complete list for coverage and live signals for speed.
  • Scoring every company and every person against seven ICP tracks before anyone wrote a word.
  • Messages written per track and per sender, so each profile sounded like itself.
  • Daily refresh of the signal flow, with signal leads sent first.
  • A sender who is really going to the event, writing to people who are going too.

Run it for your next event

The whole flow is a few prompts in Claude, ChatGPT or Cursor with AtomSell connected:

"Find posts from the past month in English, German and French where people say they are attending [EVENT NAME], and the people who comment on them. Keep the ones who fit this ICP: [YOUR ICP]. Write each a note under 200 characters that starts from their post and proposes to meet at the event."

Start a 3-day free trial with full access at app.atomsell.ai/register, connect AtomSell to Claude, ChatGPT or Cursor, and run it on your own event.

// Author

Olga Yasynska

Co-founder and GTM Lead at AtomSell. 20+ years in B2B sales and marketing, the last six building AI sales agents for LinkedIn outreach.

// Get started

Ship your first AtomSell playbook

Get an API key, connect a profile, and run a production-ready playbook end-to-end in your stack.

Create your account