Results at a glance

• 15 leads and 9 live conversations, 5 of them matching the ICP

• Four meetings booked and one call scheduled off a list built cold

• A 36.1% LinkedIn connection acceptance rate, on a 0.3% email bounce rate

A category with no industry code.

The client sells an agentic AI operating system for longevity clinics. It sits on top of a clinic's existing EMR and pulls labs, wearables and lifestyle data into one view, with AI agents handling daily health assessment, correlation analysis and member adherence. It is part of NVIDIA's Inception programme and compliant with HIPAA, GDPR, PDPA and ADHICS, which matters because its buyers sit across the USA, Europe, Singapore and the Gulf.

The commercial problem is that the category barely exists as a filterable thing. There is no industry code for a longevity clinic. A database query for wellness or preventive health in any market returns aesthetics clinics, gyms, spas and general practices, with only a thin slice of genuinely longevity-focused operators underneath. There was no outbound motion running and no evidence of which geography holds enough of those operators to be worth working.

Where they were

• No outbound running, and no tested view of which market holds the ICP

• Targeting built from broad database filters, so general wellness and aesthetics dominated the list

• A single channel, tied to one event audience

• No shared visibility, so status had to be asked for rather than read

• Zero pipeline

Four things we did differently.

• The first list was read as market information, not as a delivery problem. A broad Middle East list came back with only a small percentage meeting the ICP. Rather than re-filter the UAE, the region was deprioritised. Visible activity in Gulf wellness and healthcare is high; the density of true longevity operators is low, and filters cannot tell a clinic that runs longevity protocols from one that sells a longevity-branded package.

• An event audience used as a definition rather than an audience. Roughly 43 companies were segmented into five fit tiers, and those characteristics became the reference for everything built afterwards. An attendee list is spent in ten days; the definition keeps working.

• Map-led discovery ahead of database filters. Around 22 companies found through location and map-based research came back with visibly better fit than anything the broad filters produced, so manual discovery became the default and filters the fallback.

• Accounts validated one at a time. Each was checked with Claygent prompts before a single contact was enriched, on the standing rule that company location filters the list and contact location never does.

What changed

• The UAE tested and deprioritised, Europe carried across nine countries, a USA segment opened

• An ICP defined from real category characteristics, then applied to a cold, map-led list

• LinkedIn and email running together and measured separately

• A shared tracker pinned in Slack, daily activity visible, every pending item owned

• 9 live conversations, 5 of them ICP-qualified, 4 meetings booked

The number that matters is not nine.

Phase one ran LinkedIn against the segmented event audience and produced 11 replies and 6 leads. More usefully, it showed what a relevant prospect looks like in practice. The next event had no comparable attendee list, which ruled out repeating the approach and forced the real question: could the same profile be built from scratch?

Phase two answered it. LinkedIn booked four meetings off the cold European list, and every one of the nine conversations came from a list built and filtered from nothing, with no event giving the audience a shared reason to reply. The pipeline landed across Spain, Italy, Sweden and Poland, five sourced from LinkedIn and four from email.

The two channels did different jobs.

LinkedIn carried acceptance at 36.1%, well above normal cold benchmarks, and booked all four meetings. Email carried volume, returning 19 replies with one bounce across the entire send. Measuring them separately is what made acceptance strength and reply weakness visible instead of averaged into a single number.

The numbers

• 15 Leads generated

• 9 Active conversations

• 5 ICP-qualified

• 19 Email replies

• 36.1% LinkedIn connection acceptance

Engagement

• Client: Agentic AI operating system for longevity clinics

• Industry: Longevity and preventive health

• Markets: Europe across nine countries, after a Middle East test; a USA segment opened late

• Engagement: June to August 2026, roughly two months live across two phases

• Channels: LinkedIn · Email

• Personas: Founders · CEOs · Medical directors · Clinical and digital leaders

• Services: ICP Definition · List Building · LinkedIn Outreach · Email Outreach

• Focus: Longevity Health · ICP Discovery · EMEA