Every marketing team wants to know which channels and campaigns are driving revenue. It sounds like a simple question, but it turns out to be one of the harder analytical problems in go-to-market work. Attribution — the practice of assigning credit for a sale to the marketing and sales activities that preceded it — is genuinely complex, and most teams either oversimplify it or give up on it entirely.

This guide explains how CRM attribution works, walks through the three main attribution models, and helps you understand when your attribution data is reliable and when it’s likely telling you a misleading story.

What Attribution Actually Is

Attribution is the process of deciding how much credit each touchpoint gets for a conversion or sale. A “touchpoint” is any interaction a prospect had with your company before becoming a customer — a paid ad click, an email open, a webinar attendance, a sales demo, a free trial signup.

Most deals involve multiple touchpoints across weeks or months. The attribution question is: which of those touchpoints “caused” the sale, and how do you allocate credit across all of them?

Your CRM is central to this because it records the most important downstream events: when a lead was created, which campaigns they were associated with, when they became an opportunity, and when they closed. Without CRM data, you can measure marketing activity but not marketing outcomes.

The Three Core Attribution Models

First-Touch Attribution

First-touch attribution gives 100% of the credit for a sale to the very first touchpoint a contact had with your company.

How it works: If a prospect first encountered you through a LinkedIn ad, then attended a webinar, then had a sales conversation, then closed — the LinkedIn ad gets all the credit.

What it’s good for: Understanding which channels are most effective at generating initial awareness and starting the customer journey.

Its biggest limitation: It ignores everything that happened between the first touch and the close. If your webinar is what actually converted a lukewarm prospect into a motivated buyer, first-touch attribution never credits it.

Touchpoint	First-Touch Credit
LinkedIn ad (first touch)	100%
Webinar attendance	0%
Case study download	0%
Sales demo	0%

Last-Touch Attribution

Last-touch attribution gives 100% of the credit to the final touchpoint before conversion — usually the last marketing activity before a prospect became a lead or closed as a customer.

How it works: Using the same example, if the sales demo was the final touchpoint before the deal closed, the demo (or the channel that led to it) gets all the credit.

What it’s good for: Identifying which activities are most directly associated with conversion decisions.

Its biggest limitation: It ignores all the awareness and nurturing work that got the prospect to the point of being ready for that final touchpoint. It systematically undervalues top-of-funnel activities like content and brand advertising.

Touchpoint	Last-Touch Credit
LinkedIn ad	0%
Webinar attendance	0%
Case study download	0%
Sales demo (last touch)	100%

Multi-Touch Attribution

Multi-touch attribution distributes credit across all touchpoints that occurred in the customer journey. There are several variants:

Linear attribution gives equal credit to every touchpoint:

Touchpoint	Linear Credit (4 touches)
LinkedIn ad	25%
Webinar attendance	25%
Case study download	25%
Sales demo	25%

Time-decay attribution gives more credit to touchpoints closer to the conversion, on the logic that they had more influence on the final decision.

U-shaped (position-based) attribution splits the bulk of credit between the first and last touches, distributing a smaller share among middle touches. A common split is 40% to first touch, 40% to last touch, and 20% divided among middle touches.

W-shaped attribution adds a third significant weight — the touchpoint at which the lead became an opportunity — and distributes the remaining credit among other touches.

Model	Credit Logic	Best Used When
First-touch	100% to first interaction	Understanding top-of-funnel channel effectiveness
Last-touch	100% to final interaction	Understanding conversion triggers
Linear	Equal credit to all touches	Simple, no strong assumptions about which touch matters most
Time-decay	More credit to recent touches	Short sales cycles, where recency matters
U-shaped	Weight on first and last	Balancing awareness and conversion credit
W-shaped	Weight on first, opportunity creation, and last	Complex B2B sales with distinct lifecycle stages

How Your CRM Tracks Attribution

CRMs track attribution through a combination of lead source fields, campaign associations, and activity records. Here’s how the pieces fit together:

Lead source: When a lead is created, your CRM (or marketing automation system) stamps it with a lead source — the channel or campaign that drove the initial inquiry. This is the foundation of first-touch attribution.

Campaign associations: When a contact interacts with a marketing campaign (email, event, digital ad, content download), that interaction is recorded and associated with the contact’s record in your CRM. This is what makes multi-touch attribution possible.

Opportunity associations: When a lead becomes an opportunity, your CRM links the opportunity to the contact record, carrying over any campaign associations. The opportunity’s lead source or campaign influence fields track which marketing activities touched the deal.

Close data: When an opportunity closes (won or lost), the CRM records it, completing the chain from campaign touch to revenue outcome.

The quality of your attribution data depends entirely on how well these connections are maintained. Missing lead sources, uncleaned contact records, and incomplete campaign associations all degrade attribution accuracy.

When Attribution Data Is Reliable

CRM attribution data is most reliable when:

Your data entry is consistent. If every lead has a lead source, every contact is associated with campaigns, and your CRM fields are clean, attribution data is far more useful. Inconsistent data entry produces unreliable attribution, full stop.

Your sales cycle is short. In sales cycles under 90 days, fewer touchpoints are involved, and the path from first touch to close is more traceable. Multi-touch models are easier to trust in these conditions.

You have a high volume of deals. Attribution models are statistical. With a small number of closed deals, individual attribution numbers are too volatile to act on. You need enough closed-won deals in a reporting period for patterns to be meaningful.

You’re comparing channels at the aggregate level. Attribution is most reliable when you’re looking at patterns across many deals — “paid search consistently shows up as a first touch” — rather than trying to determine attribution for a single deal.

When Attribution Data Is Misleading

Be skeptical of your CRM attribution data when:

Your sales cycle is long and complex. A 12-month enterprise deal with dozens of touchpoints across multiple contacts is extremely difficult to attribute accurately. The buyer’s journey is too complex and too human to reduce to a model cleanly.

Your CRM data is incomplete. If a significant share of your leads have no lead source, or if contact-to-campaign associations are spotty, your attribution reports will systematically favor whichever touchpoints your team is most diligent about logging — not necessarily the ones that matter most.

You’re using attribution to make budget decisions for individual channels. Attribution models tell you about correlation, not causation. If you cut a channel because it shows low attribution credit, you may be cutting something that’s contributing to deals in ways the model isn’t capturing.

All your contacts are in one person’s name. In B2B deals where buying committees are common, attribution that tracks only one contact per account misses most of the real marketing touches. You need to track all contacts associated with the account.

Practical Recommendations for CRM Attribution

Start with one model and stick to it

The temptation is to run every attribution model and pick the one that tells the most favorable story. Resist this. Pick the model that best fits your sales cycle and business context, and use it consistently so you can build trend data over time.

For most B2B teams with multi-month sales cycles, U-shaped or W-shaped attribution is the most realistic starting point. Linear attribution is a reasonable fallback when you’re not confident in your first-touch or last-touch data quality.

Audit your lead source data quarterly

Run a report on leads with missing or “unknown” lead source fields. If more than 10-15% of your leads have no source data, your attribution reports are telling an incomplete story. Make lead source a required field in your CRM and create a short list of standardized values to prevent free-form entries that can’t be categorized later.

Use influence, not just credit

Rather than trying to determine which channel “caused” a sale, track which channels appear in the customer journey for your best customers. Campaign influence analysis — which asks “was this channel present in deals that closed?” rather than “did this channel cause the deal?” — is often more actionable and less misleading than strict attribution credit.

Connect attribution to pipeline, not just leads

Attribution models that stop at lead creation miss half the story. Push your CRM reporting to track campaign attribution through to opportunity creation and closed-won revenue. The channels that create the most leads are often different from the channels that influence the highest-value deals.

Frequently Asked Questions

Which attribution model is the most accurate? No attribution model is perfectly accurate — they’re all simplifications of a complex reality. The right model depends on your sales cycle, data quality, and what business decision you’re trying to inform. Multi-touch models (U-shaped or W-shaped) are generally more representative than single-touch models for B2B sales with multiple touchpoints, but they require better data quality to work correctly.

Does my CRM do attribution automatically? Most CRMs capture the data needed for attribution — lead source, campaign associations, opportunity records — but few automatically apply a multi-touch attribution model across your reporting. You typically need to build attribution reports yourself, or use a marketing attribution tool that integrates with your CRM to apply the model and present results. Some advanced CRM configurations include built-in attribution reports.

How do I handle offline touchpoints like events and phone calls in attribution? Offline touchpoints need to be manually logged in your CRM to be included in attribution reporting. Create standard activity types for events, phone calls, and in-person meetings, and train your team to log them consistently. If offline activities are systematically under-logged, your attribution models will underweight them — which often leads to over-investment in digital channels that are easier to track.

Can I use attribution data to justify a marketing budget request? Attribution data can support budget discussions, but it shouldn’t be the only evidence you use. Attribution models have known limitations, and experienced finance and executive stakeholders know this. Use attribution data alongside conversion metrics, pipeline influence numbers, and qualitative customer feedback about how they found you. A multi-dimensional case is more convincing — and more honest — than a single attribution report.