B2B marketing attribution: a working model

Why B2B attribution is hard

B2B purchases involve several people, many touchpoints and months of research — much of it invisible to tracking: conversations with peers, communities, podcasts, AI assistants, events. Any model that claims to credit every touch precisely is overstating what the data can do.

The goal is not perfect attribution. It is decision-grade evidence: good enough to decide where to invest next.

The three views

1. Platform view

Ad platforms and analytics tools report clicks, conversions and assisted conversions.

• Use it for: optimizing campaigns, ads, keywords and landing pages.
• Limits: each platform tends to credit itself; cross-device and offline journeys are partially invisible; conversions are not the same as revenue.

2. CRM view

The CRM records how contacts and accounts arrived, how they progressed and what they became worth.

• Use it for: judging sources by pipeline, win rate and revenue; comparing cohorts over time.
• Limits: only as reliable as source capture and lifecycle definitions; usually weighted toward first or last recorded touch.

3. Buyer view (self-reported attribution)

An open-text field on high-intent forms — “How did you hear about us?” — plus the same question asked in first sales calls.

• Use it for: surfacing influences software cannot see, especially demand creation.
• Limits: memory is imperfect and answers need light categorization.

Reconciling the views

Build a simple monthly or quarterly table by channel:

Channel	Platform conversions	CRM pipeline sourced	CRM pipeline influenced	Buyer mentions

Look for agreement and disagreement:

• All three agree — high confidence; invest with conviction.
• Buyers mention it, CRM does not — likely under-credited demand creation; protect the budget and look for better leading indicators.
• Platforms credit it, CRM does not — possible low-quality conversions; check lead quality and definitions.

Foundations to get right first

1. Source capture. Record original source reliably and prevent it being overwritten by imports and integrations.
2. Lifecycle definitions. Agree what each stage means with sales, and write it down.
3. Conversion tracking. Consistent events across the site, analytics and ad platforms.
4. Offline conversions. Send qualified outcomes back to ad platforms.

Choosing time windows

Match evaluation windows to your sales cycle. Demand capture can be judged over weeks; demand creation over quarters. Compare cohorts (for example, leads created in a quarter and what they became) rather than month-on-month snapshots.

Checklist

• Original source captured and protected in the CRM
• Lifecycle stages defined and documented
• Consistent conversion events across tools
• Self-reported attribution on high-intent forms
• A reconciliation table reviewed on a fixed cadence
• Evaluation windows that match the sales cycle

We build these foundations as part of HubSpot marketing and revenue operations.