Account-level LinkedIn attribution without clicks: what is honest to claim
Most LinkedIn ABM value shows up as impressions at accounts that later buy, not as clicks. How influenced pipeline and before/after-deal attribution work, and which claims hold up in front of a CFO.
In LinkedIn ABM, most of the effect happens without a click: people at a target account see your ads for weeks, and the deal arrives later through a demo request, a referral or an outbound reply. Account-level attribution matches LinkedIn impressions and engagement by company to CRM deals from the same company. It's the most useful view of LinkedIn's contribution, as long as you call it influence, not causation.
This article explains how account-level attribution works, the windows that make it defensible, and the line between what you can claim and what you can't.
Why click-based attribution undercounts LinkedIn
Click attribution asks: did the person who converted click an ad first? In B2B that question misses most of what LinkedIn does.
- Buying committees, not individuals. The CFO sees the ad. The finance manager books the demo two months later from a Google search. Click attribution gives LinkedIn nothing.
- Long cycles. A 90-day sales cycle is short in B2B. Cookie-based click windows and cross-device journeys break long before the deal closes.
- Low CTRs by design. A median LinkedIn single image CTR of 0.42% in one 211-company dataset (ZenABM benchmarks, 2026) means more than 99% of impressions never produce a click. If those impressions do anything, a click model can't see it.
LinkedIn itself offers view-through conversions with the Insight Tag, with a view-through window of 1, 7 or 30 days (LinkedIn Marketing API), but those still need a conversion event on your website, tied to a member. They don't connect a deal in your CRM to the company that saw your ads.
How account-level attribution works
The account is the join key. LinkedIn reports impressions, clicks and engagements by member company. Your CRM records deals against companies. Match the two by company (in practice, by website domain), and you can ask: before this deal was created, had this company seen our ads?
IntentBid does exactly this with deals from HubSpot:
- Deals come in daily from the CRM, with amount, status (open, won, lost), creation date and close date.
- Each deal is matched to its company by domain.
- A deal is influenced when its company saw or clicked your LinkedIn ads, clicked a ChatGPT ad, or came from an AI answer in the 90 days before the deal was created, or while the deal was open.
- Each deal counts once, however many campaigns touched it. Totals never double count across campaigns.
The Pipeline & revenue view then shows influenced pipeline, won revenue influenced, pipeline per $1 of spend, and for each deal which campaigns ran before it and which ran while it was open.
Before the deal vs. while it was open
These are two different claims, and separating them is what makes the report honest.
| Window | What it suggests | How to use it |
|---|---|---|
| Before the deal was created | The account was exposed while it was forming a shortlist | Evidence that LinkedIn helps create pipeline |
| While the deal was open | You kept the buying committee warm during evaluation | Evidence that LinkedIn helps close and accelerate, not create |
| Both | Ongoing presence across the cycle | Typical of well-run tier-1 programs |
If nearly all your influenced pipeline is "while open", LinkedIn isn't creating demand; it's supporting deals sales found anyway. That's still worth money, but it's a different budget argument.
What is honest to claim
Influenced pipeline is a correlation with a time order. That's more than nothing and less than proof. Here's where the line sits.
Claims that hold up
- "X of our Y new opportunities this quarter came from accounts that saw our LinkedIn ads in the 90 days before the opportunity was created."
- "Accounts that were reached had a higher rate of opportunity creation than target accounts we didn't reach." (Only if you compare like with like; see below.)
- "Our LinkedIn spend reached N of our target accounts, and M of them now have open deals."
- "Cost per engaged target account was $Z."
Claims that don't
- "LinkedIn generated $2M in pipeline." It influenced deals worth $2M. Some would have happened anyway.
- "ROAS of 12x." A ratio of influenced revenue to spend is a useful trend line, not a return. Label it as influenced.
- Adding up influenced pipeline from LinkedIn, ChatGPT Ads, events and email and presenting the total. Each channel's influence overlaps; the sum can exceed your actual pipeline.
The comparison that makes it convincing
The strongest evidence is a holdout. Take a tier of similar accounts, split it randomly, and only run ads to half for a quarter. Compare opportunity creation between the halves. It costs you reach on the held-out half, so do it on a tier where you can afford to wait, and keep the groups large enough that one deal doesn't swing the result.
Without a holdout, compare reached vs. unreached target accounts, but be careful: large accounts are both easier to reach and more likely to have deals. Compare within the same tier and size band.
Settings that make the numbers defensible
A few choices decide whether finance trusts the report.
- Use a fixed window and say what it is. 90 days before creation is a reasonable default for mid-market cycles. For enterprise deals with long research phases, the same window understates; for transactional deals, it overstates. Pick one, state it on every slide, and don't change it to make a quarter look better.
- Exclude existing customers from prospecting campaigns. Otherwise renewals and expansions show up as "influenced". IntentBid excludes companies with a won deal automatically when exclusions are on.
- Require a minimum exposure. One impression in 90 days is technically "saw our ads", but it's thin. Report both "any exposure" and a stricter version, such as 10+ impressions or one engagement.
- Keep domains clean in the CRM. Matching is by company domain. Deals attached to companies without a domain, or with a parent company's domain, won't match. Fixing this is usually the biggest single improvement to match quality.
- Remember LinkedIn's company data is approximate. LinkedIn states that professional demographic metrics, which include company, are approximate, and values under 3 events are dropped (LinkedIn Marketing API). A company with very light exposure may not appear at all. This makes influenced numbers slightly conservative, not inflated.
A reporting template that survives a CFO review
A one-page monthly view that holds up:
| Line | Example definition |
|---|---|
| Target accounts | Number on the list this period |
| Reached | Accounts with LinkedIn impressions in the period |
| Engaged | Accounts with a click, engagement, ChatGPT Ads click or website visit |
| New opportunities at target accounts | Created in the period, from the CRM |
| Of which influenced before creation | Exposure in the 90 days before |
| Influenced pipeline (before + while open) | Deal value, each deal once |
| Spend | LinkedIn + ChatGPT Ads, from the platforms |
| Pipeline per $1 (influenced) | Clearly labelled as influenced |
Add one line of commentary: what changed, and what you'll do next. Leave out channel-by-channel "ROI" claims.
Where stages fit
Attribution looks back from deals. ABM stages look forward: which accounts are moving towards a deal now. They answer different questions, and a good program uses both. Stages are covered in ABM stages sales will trust.
FAQ
Is influenced pipeline the same as view-through attribution?
No. View-through attribution credits a website conversion to an ad a person saw. Influenced pipeline matches company-level exposure to CRM deals from the same company, with no website conversion needed. It answers a B2B question, at account level.
Why 90 days?
It's long enough to cover the research phase of a typical mid-market purchase and short enough that the exposure is plausibly related to the deal. If your cycle is much longer, read the "while open" column as well; it captures exposure during long evaluations.
Can a deal count for more than one campaign?
Each campaign that touched the deal is listed against it, but in totals the deal counts once. That's the only way the totals stay smaller than your real pipeline.
What if my CRM deals have no company domain?
They can't be matched. Make sure deals are associated with a company record that has a website domain. In HubSpot, that's the company association on the deal.
Which plan includes pipeline attribution?
Every IntentBid plan includes pipeline and revenue attribution from your CRM. See pricing.