Skip to content
The honest answer is usually, “It’s complicated.” But as budgets tighten, that answer becomes harder to defend. Gartner found that marketing budgets held at 7.7% of company revenue in 2025, while 59% of CMOs said they lacked sufficient budget to execute their strategy (Gartner 2025 CMO Spend Survey).
Attribution is meant to make those budget decisions easier. The trouble is that most marketing attribution models were built around a neat sequence of trackable interactions. B2B buying rarely looks like that. It involves long sales cycles, several stakeholders, offline conversations, anonymous research, and channels that influence demand without producing the final click.
This guide explains what marketing attribution can tell you, where it becomes unreliable, and how to build a useful model without pretending it’s perfect. (New to the basics first? Our B2B marketing guide is a good place to start.)
What is marketing attribution?
Marketing attribution is the process of assigning credit for a conversion, opportunity, or sale to the marketing touchpoints associated with it.
Those touchpoints might include:
- A paid search click
- An organic article
- A webinar registration
- An email campaign
- A review-site visit
- A pricing-page session
- A demo request
Attribution modelling is the rule used to divide that credit. A first-touch model credits the earliest recorded interaction. A last-touch model credits the final one. Multi-touch models distribute credit across several interactions.
The distinction between association and causation matters. Attribution shows which recorded touchpoints appeared on a buyer’s path. On its own, it doesn’t prove that a particular channel caused the purchase or that the deal wouldn’t have happened without it. That requires incrementality testing, experiments, or other causal methods.
So the practical goal isn’t to discover one mathematically perfect version of the truth. It’s to build a consistent, decision-useful view of how marketing contributes to pipeline – and to be honest about what the data can’t show.
The main marketing attribution models
Each model answers a slightly different question. Choosing one means choosing which part of the journey you want to emphasize.
| Model | How it assigns credit | Useful when | Main limitation |
|---|---|---|---|
| First-touch | Gives all credit to the first recorded interaction | You want to understand discovery and demand creation | Ignores everything after the initial touch |
| Last-touch | Gives all credit to the final interaction before conversion | Journeys are short and conversion paths are simple | Overvalues capture channels such as branded search and demo forms |
| Linear multi-touch | Splits credit evenly across recorded interactions | You need a simple view of the whole journey | Assumes every touchpoint contributes equally |
| Time-decay | Gives more credit to interactions nearer conversion | Recent touches are likely to matter more | Can understate early education and category creation |
| U-shaped | Emphasizes first touch and lead conversion | Lead generation is the main handoff | Still centers the journey on one known lead |
| W-shaped | Emphasizes first touch, lead creation, and opportunity creation | Marketing and sales share a defined funnel | Depends heavily on clean lifecycle stages and CRM data |
No model fixes missing data. A sophisticated weighting formula applied to an incomplete journey is still an incomplete answer.
Why B2B marketing attribution is so difficult
The journey is long and crowded
An analysis by HockeyStack of 150 B2B SaaS companies reported an average of 266 measurable touchpoints and 2,879 impressions for a closed-won deal. For deals above $100,000, those figures rose to 417 touchpoints and 5,500 impressions (HockeyStack Labs, 2024).
Those are dataset-specific benchmarks, not universal constants. But they illustrate the underlying problem: a B2B decision can involve hundreds of interactions, many of which are low-intent or difficult to distinguish from genuine influence.
The buyer is an account, not one person
B2B purchases are often made by groups. One stakeholder might discover the vendor through search, another might read technical documentation, and a third might join only for commercial approval.
Most marketing attribution systems are better at tracking an individual contact than reconstructing activity across an entire account. If the opportunity contains six stakeholders but only one submits a form, the recorded journey can look much simpler than the real one.
Buyers research anonymously
Gartner found that 61% of B2B buyers prefer a buying experience without a sales representative. Buyers still value sellers for contextual questions, but they prefer self-service for many research tasks (Gartner, 2025).
That research may happen before a cookie is set, before a contact is known, across several devices, or under privacy settings that limit tracking. When someone finally requests a demo, the system may credit that last action and miss the work that created interest.
Important influence happens outside your systems
Peer recommendations, podcasts, private communities, sales conversations, forwarded documents, analyst reports, and internal meetings can all shape a purchase. Many of them produce no reliable trackable event.
This is why “not attributed” doesn’t mean “had no influence.” It may simply mean “wasn’t observable.”
Your data has boundaries
Cookie restrictions, consent choices, cross-device behavior, CRM hygiene, identity resolution, and offline activity all affect what can be measured. Attribution output should therefore be treated as a model of observed behavior – not a complete record of the buyer’s mind.
The attribution gap most B2B teams miss
The largest gap often appears before the first form fill.
A buying group may visit several pages, compare competitors, share content internally, and return repeatedly before anyone identifies themselves. Standard contact-level attribution starts when a known person enters the system, so the earlier account activity is absent.
That creates predictable distortions:
- The demo request receives too much credit.
- Branded search looks more influential than the channels that created demand.
- Early educational content appears less valuable than it is.
- The journey is attributed to one lead rather than the wider account.
- Sales and marketing work from different pictures of buyer activity.
Gartner also found that 69% of B2B buyers reported inconsistencies between a company’s website and its sellers (Gartner, 2025). Better shared visibility won’t solve that problem by itself, but it can help marketing and sales coordinate around the same evidence.
Where website visitor identification fits
Website visitor identification can add account-level context to the anonymous part of the journey by showing which companies are visiting and which pages attract their attention, subject to the tool’s coverage and the visitor’s privacy settings.
For example, it may reveal that:
- An account returned several times before submitting a form.
- Interest expanded from educational content to pricing or integration pages.
- Several sessions from the same company occurred during an open opportunity.
- A previously unknown account is showing sustained interest.
This data can strengthen marketing attribution by adding observable company-level activity that contact-only reporting misses. It can also help sales and marketing decide which accounts deserve closer attention, which is much of the same account-level thinking behind a well-run ABM program.
It should be used as one evidence source alongside CRM activity, campaign data, sales conversations, and customer research – not as a replacement for a proper marketing attribution model.
How to build a credible B2B attribution approach
1. Start with the decision
Define what attribution will inform. Budget allocation, campaign optimization, pipeline reporting, and content planning require different levels of detail. Don’t build a complicated model without a clear decision attached to it.
2. Choose a consistent conversion point
Decide whether you’re attributing leads, qualified opportunities, pipeline value, or revenue. Revenue is closest to business impact, but it also has the longest feedback loop and the messiest journey.
3. Match the model to your sales motion
Last-touch may be adequate for a short, transactional journey. A longer, committee-led sale usually needs account-level reporting and a multi-touch or stage-based model. Use the simplest model that reflects how customers actually buy.
4. Establish data rules
Document your lookback window, lifecycle stages, channel definitions, account-matching logic, and treatment of direct traffic. Without shared rules, teams can produce different answers from the same data.
5. Add account-level signals
Connect known-contact activity with observable company-level visits, opportunity data, campaign engagement, and sales interactions. Keep anonymous and identified activity distinct until you have a defensible way to connect them.
6. Compare attribution with other evidence
Use customer interviews, win-loss analysis, lift tests, geographic or audience experiments, and channel-level trends to challenge the model. If attribution says a channel has no value but buyers repeatedly name it as influential, investigate the measurement gap.
7. Report confidence, not false precision
Present attribution as directional evidence. Explain what the model includes, what it excludes, and how sensitive the result is to the chosen rules. A range with clear caveats is often more useful than an exact percentage nobody should trust.
Questions to ask before trusting an attribution report
- Does the report cover people, accounts, or both?
- What happens before the first known-contact conversion?
- Which touchpoints are unavailable or excluded?
- Are lifecycle stages applied consistently?
- Does the model reward demand creation or mainly demand capture?
- Can sales activity and offline influence be represented?
- Are anonymous visits kept separate from identified individuals?
- Would a different attribution model materially change the budget decision?
If nobody can answer those questions, the dashboard may look precise without being reliable.
FAQ
What is attribution modelling?
Attribution modelling is the method used to allocate credit across recorded touchpoints. Common models include first-touch, last-touch, linear, time-decay, U-shaped, and W-shaped attribution.
Which attribution model is best for B2B?
There’s no universal winner. The right model depends on the length of the sales cycle, the number of stakeholders, the quality of your data, and the decision you need to make. For complex sales, account-level and multi-touch views are generally more informative than single-touch reporting.
What’s the difference between attribution and incrementality?
Attribution distributes credit among observed interactions. Incrementality asks whether a marketing activity caused an outcome that wouldn’t otherwise have happened. Attribution is useful for reporting journeys; incrementality is stronger for proving causal impact.
Can website visitor identification solve B2B attribution?
It’s genuinely one of the best tools available for the anonymous research phase. Seeing which accounts are visiting your website and which channels they came in on, even before anyone fills out a form, recovers real signal that contact-only attribution misses entirely. It’s not the whole answer on its own, but for the part of the journey most attribution models can’t see at all, it’s a real, high-value fix.
Better attribution starts with clearer limits
Attribution is valuable when it helps you make better decisions. It becomes dangerous when a neat dashboard is mistaken for a complete account of why customers buy.
Choose a model that matches your sales motion. Improve the underlying data. Add account-level context where it’s defensible. Then test the story against experiments and what customers actually tell you.
If anonymous research is still missing from that picture, see how website visitor identification can add company-level context, explore the most common measurement traps in Stop Being Sabotaged by Your Attribution Strategy.
