Skip to content
Imagine
Book a demo

1/3 of Unique Engagers On LinkedIn is Your ICP

B2B content has an attribution problem, and the industry's answer has been to stop asking. Impressions went up; trust the vibes.

We think the answer is narrower and better: match the people who engage with your content against your CRM. Not modeled attribution, not brand lift - a join. This person commented on your post; this same person is a contact on an open opportunity. Count those.

The mechanics

Connect a CRM (HubSpot or Salesforce), and every engager on every post - likes, comments, shares, follows, profile views - is cross-referenced against contacts, accounts, stages, and opportunities.11Matching uses enriched engager profiles resolved to CRM contacts and accounts. Only public engagement signals are used. Three lists fall out:

  1. Engagers already in your pipeline - your content is nurturing deals you know about.
  2. Engagers matching your ICP but absent from your CRM - your content is generating names sales has never seen.
  3. Repeat engagers with no meeting booked - five interactions and no demo is a hand raised in slow motion.

What the numbers look like in practice

Across customer programs in 2026, roughly a third of unique engagers match the customer's target audience: 35% (138 of 393), 35% (35 of 101), and 28% (96 of 343).

Roughly a third of content engagers are the exact people you sell to

In one customer week, 101 people engaged, 35 matched the ideal customer profile, and 7 became marketing-qualified leads. Other programs show 28 to 35 percent target-audience shares.

ONE WEEK, THREE POSTS (MONTH 3)101 people engaged35 matched the ICP35% of engagers7 MQLsCEOs and CTOs of YC-backed AI startups, plus an enterprise Chief Cloud OfficerTARGET-AUDIENCE SHARE OF UNIQUE ENGAGERSCustomer A35.1% · 138 of 393Customer B34.7% · 35 of 101Customer C28% · 96 of 343

Imagine AI customer programs, 2026 · engager lists matched against each customer’s ICP definition · MQL = the customer’s qualification bar

One representative week for a four-person YC company selling to CTOs: three posts produced 101 engagers, 35 ICP matches, and 7 MQLs - including CEOs and CTOs of YC-backed AI startups and a Chief Cloud Officer at an enterprise infrastructure company. The weekly output of the program is not a chart; it is this list.

And when the CRM join runs against revenue:

When engagement is matched against a CRM, revenue shows up

One program connected 82 percent of matchable CRM records to content. Another tied 290 thousand dollars of 2024 revenue to engaged accounts. Imagine AI's CRM maps roughly 970 thousand dollars of closed annual recurring revenue to LinkedIn engagement.

82%of one customer’s matchable CRM records(70 of 85) had a direct buyer or same-companyconnection to their contentONE CUSTOMER, APR–JUL 2026$290Kof one customer’s 2024 revenue tied toaccounts that engaged with content(350K CRM contacts matched)CRM-TO-GRAPH MATCH, 2026~$970Kclosed ARR mapped back to LinkedInengagement on our own CRMIMAGINE AI INTERNAL, 2026We count only engagement we can see and map — comments, likes, shares, follows, and profile views. Never reads, saves, or forwards.

Imagine AI customer data, 2026 · engager-to-CRM mapping (HubSpot / Salesforce) · figures are per-customer programs, not averages

For one customer, 82% of matchable CRM records (70 of 85) had a direct buyer or same-company connection to their content. For another, matching 350K CRM contacts against the engagement graph tied $290K of 2024 revenue to accounts that engaged. On our own CRM, we map roughly $970K of closed ARR back to LinkedIn engagement.22These are program-specific figures rather than averages; each reflects one customer's CRM and measurement window.

Scale changes the texture, not the pattern. Across five Rippling executives over six months - 94 posts, 515K impressions, a 2.45% engagement rate (about 2-3x the B2B benchmark) - the engager pool contained 453 distinct external ICP profiles: 113 founders and 341 senior buyers (CFOs, CMOs, CHROs, VPs of Finance and People) at companies like Visa, Meta, Google and Nutanix.33Six months across five tracked executive accounts, deduplicated to unique people. About 20–29% of engagement was external target audience; the rest was internal.

The honest boundary

This method only counts what it can see: public engagement mapped to a CRM. It cannot see reads, saves, or a post being forwarded in a Slack channel - which means these numbers are floors, not ceilings. We prefer an undercount we can defend to a model we can't.

Two corollaries worth internalizing:

  • Impressions are an input, not an outcome. A 400K-impression week matters if, and only if, the engager list underneath it contains buyers.
  • The baseline comes first. Before changing anything, run the join on your existing content history. Most teams discover their past content already touched pipeline - it just was never counted.

The monthly loop

Month 0 establishes the baseline: historical performance, the engager list, the CRM overlap. Every month after is measured against it - engagers, ICP matches, CRM-matched accounts, and the flagged repeat-engagers your sales team should call. Content stops being a faith-based program and becomes a list your revenue team argues about. That is the point.

Engager-to-CRM mapping ships with native HubSpot and Salesforce apps. The same join powers per-post CRM metrics: which posts touched pipeline, which touched closed-won.

More from Research