ABM attribution and why account-based programs break the model

Attribution models built for lead funnels can't score an account-based program, and changing the weights won't fix them. They credit a converted lead. ABM works on a buying committee inside an account you chose before anyone showed interest, touched by ads, outbound, events, and content in the same few weeks, with pipeline arriving a quarter later. The unit the model credits and the unit the program works on are different things.

That's why account based marketing attribution turns into an argument every quarter. Here's where each model breaks, and what to put in front of finance instead.

TL;DR

• Last-touch credits whoever arrived last, which in ABM is usually the demo request. The program's work happened before that.
• Multi-touch splits credit by a rule nobody agreed to, across channels that hit the same account in the same week.
• Account-level rollups like MQAs lose the person, so the number can't be checked.
• Replace the model with a record: observed touches per named contact on the deals that opened, plus a comparison with accounts you didn't reach.
• Frame it as influence, not credit. The question is what marketing touched on the way to the deal, not whose deal it was.

Two different units of measurement

A lead funnel has a natural anchor: the form fill. Someone converts, a record is created, and attribution works backward from that moment. We've written about how much of the journey that anchor misses(opens in a new tab) even in a demand gen program. In ABM the anchor is in the wrong place entirely.

You picked the account before it did anything. You identified the people on the buying committee, often six or more of them on a closed-won enterprise deal, and you reached them through several channels at once. Most of them never fill out a form. The one who does is typically the champion, late, after the committee has already been working on the decision for months. A model anchored to that conversion sees the last few metres of a long journey and credits whatever happened to be standing there.

So before you choose between last-touch and multi-touch, notice what they share. Both treat a person converting as the event to explain. ABM's event is a committee moving, and no amount of weighting turns one into the other.

Where last-touch breaks

Last-touch gives the whole deal to the final interaction before the opportunity opened. In an ABM program that interaction is almost always the demo request or the sales reply, because that's what opportunities are created from.

Everything the program did to get there disappears: the ads the committee saw for two months, the pricing-page visits from three different people, the event the champion attended. Worse, last-touch rewards the channel closest to the conversion, so the rational response is to spend more on branded search and retargeting and less on the work that created the demand. Teams end up defunding the part of the program that worked because the model can't see it working.

Where multi-touch breaks

Multi-touch tries to fix that by spreading credit across every tracked interaction. For a lead funnel with a handful of channels and a short cycle, it's a reasonable approximation. For ABM it fails on three fronts at once.

The weights are a rule, not a finding. Even-weighted, time-decay, W-shaped: each is a decision about how credit should be split, made before the data arrives. Two reasonable people will pick different rules and get different answers from the same deal, and the rule, not the campaign, decides which channel looks good.

The touches aren't independent. ABM deliberately surrounds an account. Ads, direct mail, outbound, and an event hit the same people in the same fortnight because the combination is the play. Splitting credit between them treats a coordinated motion as a set of competing channels.

The lag swamps the signal. Pipeline from a target account can open two quarters after the touches that mattered. By then the campaign has ended, budget has moved, and the model is attributing this quarter's deals to last spring's spend. A dashboard that updates monthly and a program that pays back in six months were never going to agree.

There's a practical audit worth running before trusting any multi-touch report: check whether your UTMs record the origin of a journey or just the last touch, whether the model is even-weighting everything by default because nobody set it up, and whether your opportunities have contact roles filled in at all. A model can't credit a committee member who was never attached to the deal.

Where account-level rollups break

Account-based platforms answered the attribution problem with account-level metrics: marketing-qualified accounts, engaged accounts, influenced accounts. These solve the lag problem by moving earlier in the journey, and they create a new one. "We influenced Acme" can't be checked. Which people? What did they do? Did any of them have anything to do with the deal?

The detail was rolled up before it reached the report, which is also why account scores go wrong(opens in a new tab) as evidence: the person behind the signal disappears. An MQA threshold says an account crossed a line. At renewal, when finance asks what those accounts became, nobody can say, and the program measured on them gets cut.

What to report instead

Replace the model with a record. Three pieces.

1. 1
   Observed touches per named contact, on the deals that opened. For each opportunity, which members of the buying committee did marketing reach, through which programs, on which dates. No weights. The report says four of the six people on this deal saw your ads and two visited pricing in the ninety days before the opportunity opened, and anyone can click through to the names.
2. 2
   Touched against untouched. Compare meetings, pipeline, and win rate for accounts the program reached with similar accounts it didn't. The gap is the program's effect. Our ABM playbook(opens in a new tab) measures the whole program this way rather than trying to isolate channels, because inside a surround-sound program the channel question never settles.
3. 3
   Coverage, not credit. Report how much of each buying committee you reached before the deal opened. Coverage is a number marketing controls, it's checkable, and it's the mechanism by which the program influences the deal in the first place.

This is the shift from attribution to influence. Attribution asks who deserves the deal, and sales will always have a view on that. Influence asks a question with an observable answer: what did marketing touch on the way to it? That's the framing that survives a CFO conversation.

It also changes what you can defend. The ABM program at Fingerprint(opens in a new tab) moved its tracking from the account to the named contact, and the team described the result as enterprise ad spend that became defensible for the first time: every enterprise win could be traced back to a first-party signal from an identified person.

What it looks like on a real deal

Take a hypothetical opportunity that opened in June. The record shows a VP of Operations who clicked a LinkedIn ad in March and visited the integrations page twice in April. A director on the same account joined a webinar in May. A third contact, in finance, saw the ads and never clicked. Sales first spoke to the account in late May, and the opportunity opened three weeks later.

A last-touch model credits the sales conversation. A multi-touch model gives the webinar a share it never earned and the March ad a share nobody can defend. The record simply says marketing reached three of the five people on the committee before sales did, names them, and dates each touch. Nobody has to agree on percentages, and the person who asks "which people?" gets an answer.

How Vector handles ABM attribution

Vector keeps ad engagement, site visits, meetings, and the deal on one timeline per account, with people attached only when a real person has been confidently identified. When an opportunity opens or closes, the account's story lists the marketing moments on the way there, by contact and by date.

This is deliberately not multi-touch attribution. There's no model carving a deal into percentages. Each story shows the real touches in order, which ads reached which people, and where sales picked it up. If a champion changes jobs, their engagement stays credited to the account where it happened, so the story holds up months later.

For the cohort view, the same timeline lets you compare deals your ads touched with deals they didn't. Reporting is included in every plan, so the proof arrives with the campaign rather than as a separate purchase.

If your attribution report and your pipeline disagree every quarter, see how Vector handles ABM(opens in a new tab).

Book a demo →(opens in a new tab)

Frequently asked questions

What is account based marketing attribution?

It's the attempt to connect ABM program activity to pipeline and revenue from the target accounts. The difficulty is that ABM works on a buying committee inside a chosen account across many channels at once, while most attribution models are built to credit a single converted lead.

Why doesn't multi-touch attribution work for ABM?

Three reasons. The weights are a rule chosen before the data arrives, so the rule decides which channel looks good. ABM touches are coordinated rather than independent, so splitting credit between them misreads the play. And pipeline opens a quarter or more after the touches that mattered, so the model credits this quarter's deals to spend that has already moved.

What should you measure instead of ABM attribution?

Observed touches per named contact on the deals that opened, with programs and dates attached; a comparison of meetings, pipeline, and win rate between accounts the program reached and similar accounts it didn't; and buying-committee coverage before the deal opened.

What is the difference between attribution and influence in ABM?

Attribution asks which channel or team deserves credit for the deal. Influence asks what marketing touched on the way to it, by person and by date. Influence is observable and checkable, which is why it holds up better with finance.

Is an MQA a form of attribution?

Not really. A marketing-qualified account says an account crossed a scoring threshold. It doesn't say which people engaged or whether any of them were on the deal, so it can't be traced back to pipeline when finance asks.