Attribution Model Framework

Use this framework when setting up or revising your marketing attribution model. Attribution answers the question “which marketing activities are creating pipeline and revenue?” — and the answer changes significantly depending on which model you choose. This framework helps you pick the right model, implement it, and avoid the most common mistakes.

Attribution Model Comparison

There is no universally correct attribution model. Each one tells a different story about your marketing performance.

Model	How Credit is Assigned	Best For	Blind Spots
First Touch	100% to the first interaction	Understanding demand generation and top-of-funnel performance	Ignores everything that happens after first touch
Last Touch	100% to the interaction before conversion	Understanding what triggers conversions	Ignores all nurture and awareness activity
Linear	Equal credit to every touchpoint	Simple multi-touch view when you lack data to weight touches	Treats a casual blog visit the same as a demo request
Time Decay	More credit to touches closer to conversion	B2B sales cycles where recent activity matters more	Undervalues early awareness that started the journey
U-Shaped (Position)	40% first touch, 40% lead creation, 20% distributed across middle	Balancing demand gen and conversion credit	Ignores sales-assisted touches post-MQL
W-Shaped	30% first touch, 30% lead creation, 30% opportunity creation, 10% middle	Full-funnel B2B with clear stage transitions	Complex to implement; requires clean stage data
Custom / Data-Driven	ML model assigns credit based on statistical impact	Mature teams with 12+ months of conversion data	Black box; hard to explain to stakeholders

Choosing Your Model

Answer these questions to determine the right starting model:

Question	Answer	Recommended Model
Do you have fewer than 200 closed-won deals?	Yes	Start with First Touch + Last Touch (run both)
Is your sales cycle under 30 days?	Yes	Last Touch or Linear
Is your sales cycle 30-90 days?	Yes	U-Shaped
Is your sales cycle 90+ days with 10+ touchpoints?	Yes	W-Shaped or Time Decay
Do you have a data science team?	Yes	Custom / Data-Driven (after implementing a rules-based model first)

General recommendation for B2B SaaS: Start with U-Shaped attribution. It gives credit to both the awareness activity (first touch) and the conversion activity (lead creation) while distributing some credit to nurture touches. Move to W-Shaped once your CRM has clean opportunity creation data.

Implementation Steps

Step 1: Define Your Touchpoints

List every interaction you can track. Be honest about what you actually measure vs. what you wish you measured.

Touchpoint	Channel	Trackable?	How Tracked
Organic search visit	SEO	Yes	Google Analytics UTM + landing page
Paid ad click	Paid	Yes	UTM parameters + ad platform
Blog post read	Content	Yes	GA page views + session data
Email click	Email	Yes	MAP click tracking
Webinar attendance	Events	Yes	Webinar platform integration
SDR cold call	Outbound	Partially	CRM activity logging (depends on rep discipline)
Conference conversation	Events	No	Manual entry if at all
Word of mouth	Referral	No	Self-reported in “How did you hear about us?”
LinkedIn organic post	Social	Partially	UTM links on some posts, dark social on others

The trackability column matters. Your attribution model is only as good as the data feeding it. Invest in tracking infrastructure before investing in model sophistication.

Step 2: Map Touchpoints to CRM Stages

Define which touchpoints map to each attribution stage:

CRM Stage	Attribution Moment	Example Touchpoints
Anonymous → Known	First Touch	First website visit, first ad click
Known → MQL	Lead Creation	Form fill, demo request, content download
MQL → SQL	Sales Accepted	SDR outreach, discovery call booked
SQL → Opportunity	Opportunity Creation	AE creates opp after qualified discovery
Opportunity → Closed Won	Revenue	Contract signed

Step 3: Set Up Tracking Infrastructure

• UTM parameter convention documented and enforced across all channels
• CRM fields created: First Touch Source, First Touch Medium, First Touch Campaign, Lead Creation Source
• Marketing automation platform (HubSpot, Marketo) syncing touchpoint data to CRM
• Website analytics tracking all form submissions with source attribution
• Offline events (conferences, dinners) have a manual entry process that sales follows
• “How did you hear about us?” field on key forms (supplement, not replace, digital tracking)
• Data warehouse receiving CRM + MAP + analytics data for multi-touch analysis

Step 4: Build Attribution Reports

Report 1: Pipeline by First Touch Source

First Touch Source	Pipeline Created ($)	Pipeline Created (#)	Win Rate	Revenue ($)
Organic Search
Paid Search
Paid Social
Outbound (SDR)
Referral / Partner
Direct / Unknown
Events

Report 2: Pipeline by Lead Creation Source

Lead Creation Source	MQLs	SQLs	Opportunities	Revenue ($)	CAC
Demo Request
Content Download
Webinar Registration
Free Trial Signup
SDR Outbound
Inbound Call

Report 3: Multi-Touch Attribution (if using U-Shaped or W-Shaped)

Campaign / Channel	First Touch Credit ($)	Lead Creation Credit ($)	Opp Creation Credit ($)	Middle Touch Credit ($)	Total Attributed ($)
{{campaign}}

Build these reports in your analytics platform and include the top-line numbers in your weekly GTM report.

Step 5: Validate and Calibrate

Attribution models are wrong by definition — they are simplifications of complex buying journeys. Validate quarterly:

• Compare attributed revenue total to actual closed-won revenue (should be within 5% for single-touch, exact for multi-touch)
• Review 10 recent closed-won deals manually: does the attributed source match what the AE says actually drove the deal?
• Check for “dark funnel” blind spots: are high-performing channels like podcasts, communities, or word-of-mouth being captured at all?
• Compare model outputs month-over-month: sudden swings usually indicate tracking problems, not real performance changes

Common Attribution Pitfalls

Pitfall	Why It Happens	How to Fix
”Direct/Unknown” is your #1 source	UTM parameters not applied consistently	Audit every channel for UTM coverage. Add mandatory UTM fields to campaign launch checklists.
SDR outbound gets zero multi-touch credit	Outbound calls and emails are logged in CRM but not connected to the attribution system	Map CRM activities as touchpoints in your attribution tool.
Self-reported source contradicts digital attribution	Buyer says “I found you through a podcast” but first touch shows organic search	Track both. Use self-reported as a qualitative signal alongside digital attribution.
Attribution changes every month	Small sample sizes cause volatility	Do not make budget decisions on fewer than 20 conversions per channel per month.

How to Customize

• For companies spending under $50K/month on marketing, start with first-touch and last-touch attribution running in parallel. The additional complexity of multi-touch models is not justified until you have enough data volume and enough channels to warrant it. Use the savings in complexity to improve tracking quality instead.
• For companies with strong outbound motions, add a dedicated “Outbound Sourced” vs. “Outbound Assisted” split. An opportunity sourced by an SDR cold call is different from one where marketing generated the lead and the SDR followed up. Both are valuable, but they should be measured separately. Reference your SDR operations setup for how to track this.
• For ABM-focused teams, shift from lead-level attribution to account-level attribution. Sum all touchpoints across all contacts at a target account and attribute at the account level. This prevents the common problem where ABM campaigns show poor lead-level ROI because the buying committee has 5-8 people, each with a few touches, that add up to a strong account-level signal. See the ABM use case for more detail.