Account-based attribution: how to measure ABM from first touch to pipeline

Key takeaways

• Account-based attribution credits every person and every touchpoint at a target account, instead of the 1 person who filled in the form. In B2B, the account is the unit that buys.
• Lead attribution undercounts the channels that reach a buying committee. One team's 9 reported leads were 87 at company level, and about 50 of them came from LinkedIn without a form fill.
• Most LinkedIn engagement never becomes a click. Measured at the account level, about 50% of closed revenue at one company involved accounts its ads had already reached.
• Pipeline visibility means 4 views: which accounts are in market, which channels created pipeline, where accounts stall, and how long first touch to deal takes.
• The setup is the hard part: group people into companies, identify the anonymous visitors, keep the CRM clean, and judge channels on pipeline and revenue, not leads.

What is account-based attribution?

Account-based attribution gives the credit for a deal to the marketing touchpoints of everyone at the account that bought, not just the 1 person who converted. The unit of measurement is the company, because in B2B the company is what buys.

Lead attribution treats every contact as a separate story. A marketing manager clicks a LinkedIn ad in January. A head of sales reads a comparison page from a Google search in March. A CFO books a demo from a brand search in May. Lead attribution sees 3 unrelated people, credits the brand search with the deal, and calls the LinkedIn ad a lead that never converted.

Account-based attribution sees 1 account with 1 journey and 3 people in it. The LinkedIn ad started it, the comparison page moved it, the brand search closed it. All 3 get credit, and the budget conversation changes.

Why account-based marketing needs it

ABM picks the accounts first and runs campaigns at them. That breaks lead attribution in 2 ways. First, the people your ads reach are rarely the people who fill in the form. Second, most of the engagement never becomes a click: buying committees research a vendor on LinkedIn without clicking anything, and some of the best accounts never visit the website at all.

So an ABM program measured on form fills looks like it's failing while the pipeline it created lands under "direct" or "brand search". Account-based analytics software exists to close that gap: it rolls people up into accounts, counts exposure as well as clicks, and follows the account into the CRM pipeline.

How does account-based attribution work?

It's multi-touch attribution with 1 extra step: before the credit is split, every person is rolled up into their company. (The multi-touch basics, if you want them first.)

1. Company identity. Every visitor is tied to a company 1 of 3 ways. Known people are grouped by the company on their form, signup, or CRM record. Anonymous visitors are matched by IP to a company with Lens, so a pricing-page visit from an office counts before anyone fills in a form. And LinkedIn reports ad impressions and engagement per company, so accounts that saw your ads and never clicked still show up.
2. The account journey. Every touchpoint from every person at the company, on 1 timeline: ad impressions, clicks, page views, form fills, then the CRM events (contact created, deal created, stage changes, deal won). The journey runs up to the moment the account becomes pipeline. After that, sales owns it.
3. The model. The same 4 rules as person-level attribution, applied to the account's journey: first touch, last touch, linear, U-shaped. A U-shaped split now credits the person who opened the account and the person who converted it, who are usually not the same person.
4. The revenue join. CRM deals and payments are matched to the account's journey. Because deals are matched by company, a deal signed by someone who never visited the website still lands on the touchpoints of their colleagues who did.

Two things decide whether the numbers hold up.

Grouping discipline. Company attribution only works if every tracking method uses the same company ID. Turn on company attribution in workspace settings, call group() wherever you identify someone, and keep 1 record per company in the CRM. Two CRM records for the same account mean 2 half-journeys.

Counting exposure, not only clicks. LinkedIn's click-through rates underrepresent its influence. If your tool only counts clicks, the channel that reaches buying committees looks like it does nothing. Spectacle imports LinkedIn impressions, video views, and document reads per company and puts them on the account timeline next to the CRM stages.

Lead attribution vs account attribution, on 1 deal

Here's 1 target account. 3 people, 6 touchpoints, 1 deal of €36,000. The numbers are illustrative, the shape is typical.

1. Week 1. The marketing manager sees a LinkedIn thought-leader ad. No click.
2. Week 2. The marketing manager clicks a LinkedIn ad and reads a blog post.
3. Week 5. The head of sales arrives from a Google search and reads a comparison page.
4. Week 6. The head of sales comes back direct and visits pricing twice.
5. Week 9. The CFO searches your brand on Google, clicks the ad, and books a demo.
6. Week 14. Deal won: €36,000.

What lead attribution reports

The CFO is the only lead. Her journey is 1 visit long. Google brand search gets €36,000, LinkedIn gets €0, and the comparison page never appears. The marketing manager shows up as a LinkedIn lead that never converted.

What account attribution reports

U-shaped: 40% to the first touch (the LinkedIn ad view), 40% to the demo (Google brand), 20% spread over the 3 visits in between. Linear: 5 visits at €7,200 each.

Two things changed. The ad view counts: the person who started the journey never clicked. And the direct pricing visits join the account instead of being a mystery, because Lens tied the office IP to the company.

Same deal, and LinkedIn goes from €0 to €16,800 before anyone changes a model. The lesson for ABM: the unit of measurement matters more than the model.

Pipeline visibility: the 4 views a marketing manager needs

Attribution answers "which channels made this happen?" Pipeline visibility answers "what's happening right now, and where?" ABM needs both. These are the 4 views, with the question each answers and where the number comes from.

More on stage timing in Audience insights.

Two numbers that belong on the Monday dashboard: the share of new deals whose contact was already in the audience for over 90 days, which tells you how far ahead marketing works, and journey coverage, the share of new deals with a tracked journey at all. If coverage is low, fix tracking before reading anything else.

What changes when you measure accounts instead of leads

3 situations from Spectacle accounts, anonymized, where the account view changed a budget decision.

The LinkedIn budget that almost got cut

In the lead report, LinkedIn produced 9 leads. At company level the same campaigns touched 87 accounts, and about 50 of them had engaged on LinkedIn without ever filling in a form. At another B2B software company, LinkedIn influenced about 43% of pipeline and about half of 12-month revenue. Neither number exists in a lead report.

Action: judge LinkedIn on account-level pipeline. Budget it against the deals its accounts became, not against form fills. How company-level LinkedIn tracking works.

The deal that started with 1 post

A $16M deal traced back to 1 organic LinkedIn post, 4 months before the first form fill. The person who engaged with the post was not the person who booked the demo.

Action: keep every touchpoint for as long as the cycle runs, with no lookback window. A 90-day window loses this deal's first touch 3 times over.

The HubSpot funnel that overcounted

HubSpot showed about 46 contacts reaching the deal-won stage. Grouped by company, it was 9 businesses. Lead-level reporting made the conversion rate look 5 times better than it was, and hid that Google Ads sparked the first touch while direct, referral, Bing, and LinkedIn carried the deal to close.

Action: report funnels per company, and show the sources that influenced every stage, not only the first or the last. (HubSpot revenue attribution reporting.)

Common account-based attribution problems (and how to fix them)

1. People never grouped into companies. The most common one. Without group(), 3 colleagues are 3 partial journeys and the account view is empty. Call group() wherever you identify someone, and keep the group ID identical across the snippet, GTM, Segment, and the REST API.
2. Two CRM records for 1 account. The account journey splits in 2 and both halves look short. Keep 1 company record per account, 1 currency and 1 amount field per deal, and real closed-won dates.
3. Clicks counted, exposure ignored. An ABM program that runs thought-leader and awareness ads gets no credit on click-only measurement. Connect LinkedIn so impressions and engagement land on the account timeline.
4. Target accounts that never visit the site. Some of the best accounts engage only on LinkedIn. Put the account list in Spectacle, sync it to LinkedIn as a matched audience, and report on the accounts the ads reached that later became pipeline. LinkedIn needs at least 300 matched companies before it delivers against a list.
5. Anonymous office traffic filed as direct. Repeated pricing visits from 1 company are a shortlist signal, not direct traffic. Turn on Lens, and route tracking through your own domain with server-side tracking so corporate firewalls and ad blockers don't drop the visits.
6. A lookback window shorter than the deal cycle. The gap between first ad and CRM record can run months. GA4 stops at 90 days, Meta at 7. Use a tool with no lookback cap.
7. Reading influence as cause. "50% of closed revenue involved accounts the ads reached" shows presence, not lift. To prove the program caused the pipeline, hold out a slice of the target list and compare.

How to set up account-based attribution in 5 steps

Most teams are done in about 30 minutes. Steps 1 and 3 decide whether the account view is trustworthy.

1. Clean the CRM, 1 record per account

1 company record per account, 1 currency and 1 amount field per deal, a defined lead stage, source fields nobody overwrites, and real closed-won dates. Every account-based tool reports garbage on duplicate accounts.

2. Track every page, on your own domain

Add the JavaScript snippet to every page, or use the Google Tag Manager template, Segment, or the REST API. Route it through server-side tracking so enterprise visitors behind firewalls and ad blockers still show up.

3. Group people into companies and turn on company attribution

Call identify() whenever someone shares contact details, and group() with the company ID. Then switch on company attribution in workspace settings. From here on, journeys roll up per account.

4. Add the people who never convert

Connect LinkedIn Ads so impressions and engagement land per company, with 24 months of history on day 1. Turn on Lens so anonymous office visits are named. Put your target account list in Spectacle and sync it to LinkedIn as a matched audience.

5. Connect revenue and ad spend, then report per account

Connect your CRM (HubSpot, Salesforce, Dynamics 365, Odoo) and ad accounts so pipeline and spend import on their own. Build 1 dashboard with the 4 views above and put it in the weekly pipeline meeting, where sales already sits. Adoption is scheduled, not assumed.

Choosing account-based analytics software: 6 things to check

Most tools in this category show you which accounts are on your site. Fewer follow those accounts into the CRM and back to the campaign that reached them. Check these before you buy.

Price is the other check. Enterprise ABM platforms start around €1,000 a month and go to €50,000 a year with modules. Spectacle's account-level attribution, LinkedIn company attribution, Lens, and audience syncs are on the €349 Scale plan. See pricing.

Account-based attribution FAQ

What's the difference between lead attribution and account-based attribution?

Lead attribution credits the touchpoints of the 1 person who converted. Account-based attribution credits the touchpoints of everyone at the company that bought, including people who never filled in a form and visits that happened before anyone was identified.

Which attribution model works best for ABM?

U-shaped, applied at the account level. It credits the touchpoint that opened the account and the one that turned it into pipeline, which are usually 2 different people. Check it against linear: an account-level channel that ranks high in both is safe to scale.

Can you attribute LinkedIn ads without clicks?

Yes. LinkedIn reports impressions and engagement per company. Matched to the account journey, an ad view counts as a touchpoint, so a deal whose first touch was a thought-leader ad nobody clicked still credits LinkedIn.

How do you identify the companies behind anonymous visits?

By matching the visitor's IP address to a company record. Lens does this in real time and returns the company name, industry, size, and location. It identifies companies, not people, and skips visitors who are already identified.

What does pipeline visibility mean in ABM?

Seeing, per target account, which ones are researching you now, which channels created the pipeline, where deals stall, and how long first touch to deal takes. Attribution tells you what worked; pipeline visibility tells you what's happening.

How much does account-based analytics software cost?

Enterprise ABM platforms start around €1,000 a month. Spectacle's account-level attribution, LinkedIn company attribution, Lens, and ABM audience syncs are on the €349 a month Scale plan, with a 14-day free trial.

See every account behind your pipeline

Your lead report credits 1 person for a deal that 3 people and 6 touchpoints made. Spectacle rolls every person up into their account, counts the LinkedIn ads they saw and the office visits they never announced, and follows the account into your CRM pipeline on all 4 models. It starts at €349 a month on the Scale plan and takes about 30 minutes to set up.

Start a free 14-day trial or book a demo.

Related reading

• Multi-touch attribution: what it is, 4 models, and how to set it up
• Conversion funnel tracking and audience segmentation
• HubSpot revenue attribution reporting
• Cohort-aligned revenue: a different attribution model
• Revenue attribution analysis explained
• The 7 best marketing attribution tools for SaaS
• LinkedIn company attribution and company identification (Lens)