The Full-Funnel Attribution Model: Who Deserves Credit?

A customer’s journey:

Monday: Sees your Facebook ad. Clicks it. Bounces.
Tuesday: Searches for your brand on Google. Clicks paid search. Browses pricing page.
Wednesday: Receives a retargeting email. Clicks. Signs up for demo.
Thursday: Books a demo via your website (no click from ad).
Friday: Completes demo. Buys.

Who deserves credit for the sale?

Facebook? (They started the journey)
Google? (They brought them back)
Email? (They nudged them over the finish line)
Your website? (They closed it)

The answer depends on which attribution model you use. And that answer changes your entire strategy.

The Four Attribution Models

Model 1: First-Click (First Touch)

Give 100% of the credit to the first interaction.

In the example above: Facebook gets all the credit.

Facebook drove the awareness. Without Facebook, the customer never would have searched for you on Google.

Assumption: The first touchpoint is most important because it creates awareness.

Model 2: Last-Click (Last Touch)

Give 100% of the credit to the last interaction.

In the example above: Your website gets all the credit (no ad channel involved in the final step).

Actually, if we count email as the “last click before conversion,” email gets the credit.

Assumption: The last touchpoint is most important because it closes the deal.

Model 3: Linear (Equal Weight)

Give equal credit to every touchpoint.

In the example above:

Facebook: 25% (1 of 4 interactions)
Google: 25% (1 of 4 interactions)
Email: 25% (1 of 4 interactions)
Website: 25% (1 of 4 interactions)

Assumption: All touchpoints are equally important. None would have worked without the others.

Model 4: Time-Decay (Recency-Weighted)

Give more credit to recent interactions, less to early ones.

Example (using exponential decay):

Facebook (Monday): 10% (oldest)
Google (Tuesday): 20%
Email (Wednesday): 30%
Website (Thursday): 40% (most recent)

Assumption: Recent touchpoints are more important because they’re closer to the decision. But early touchpoints matter too for awareness.

Why Attribution Matters

Let’s see how these models change your spending decisions.

You have 100 customers with similar journeys to the example above. You spent:

Facebook: $10,000
Google: $8,000
Email: $3,000
Website: $0
Total marketing spend: $21,000

Attribution by First-Click:

Facebook: 100 customers → ROI = $100,000 revenue / $10,000 spend = 10x
Google: 0 customers → ROI = N/A (not getting credit for any conversions)
Email: 0 customers → ROI = N/A
Conclusion: “Facebook is crushing it. Let’s double Facebook spend and cut Google and email.”

Attribution by Last-Click:

Facebook: 0 customers → ROI = N/A
Google: 0 customers → ROI = N/A
Email: 50 customers → ROI = $50,000 revenue / $3,000 spend = 16.7x
Website: 50 customers → ROI = N/A (free)
Conclusion: “Email is our best performer. Double email spend. Cut Facebook and Google.”

Attribution by Linear:

Facebook: 25 customers → ROI = $25,000 revenue / $10,000 spend = 2.5x
Google: 25 customers → ROI = $25,000 revenue / $8,000 spend = 3.1x
Email: 25 customers → ROI = $25,000 revenue / $3,000 spend = 8.3x
Website: 25 customers → ROI = $25,000 revenue / $0 spend = infinite
Conclusion: “Email and website are best performers. Allocate budget proportionally.”

Attribution by Time-Decay:

Facebook: 10 customers → ROI = $10,000 / $10,000 = 1x
Google: 20 customers → ROI = $20,000 / $8,000 = 2.5x
Email: 30 customers → ROI = $30,000 / $3,000 = 10x
Website: 40 customers → ROI = $40,000 / $0 = infinite
Conclusion: Similar to linear, but weighted toward recent touchpoints.

Same customer journey. Different model. Completely different strategy.

Which model is right?

All of them. And none of them.

The Problem: There Is No Perfect Model

Each model has flaws:

First-Click is wrong because:

If you cut Google and email, customers won’t convert even if they see the Facebook ad
You’re overvaluing awareness and undervaluing conversion
Leads to overspending on top-of-funnel

Last-Click is wrong because:

If you cut Facebook, customers never search for you
You’re overvaluing conversion and undervaluing awareness
Leads to underspending on awareness

Linear is wrong because:

All touchpoints are NOT equally important
Early awareness is less valuable than late-stage conversion
Oversimplifies the customer journey

Time-Decay is better but still wrong because:

It’s arbitrary how much to weight recency
It ignores the actual business impact of each channel
Different channels have different roles; time alone doesn’t determine value

How to Actually Think About Attribution

Instead of picking a model and trusting it, use multiple models and triangulate.

Step 1: Calculate all four models

See how they rank your channels. Usually they’ll agree on some things and disagree on others.

If all four models agree that “Email is best performing,” it’s probably true.

If only one model says “Facebook is best,” it’s probably wrong.

Step 2: Understand the role of each channel

Facebook/Social: Awareness. Job is to get people interested.
Google/Search: Consideration. Job is to bring back interested people.
Email/Retargeting: Conversion. Job is to close the deal.
Website: Infrastructure. It facilitates all conversions.

Once you understand roles, you can assign credit more thoughtfully.

If Facebook’s job is awareness and Google’s job is consideration, they deserve different credit even if they have the same “touchpoints.”

Step 3: Run incrementality tests on your top channels

The truth about attribution is in incremental testing (see “Incrementality Testing” in the Forecasting section).

Instead of arguing about which model is right, run an experiment:

Turn off Facebook for a week
Measure total conversions
Do conversions drop by 25% (first-click estimate) or 10% (time-decay estimate)?
That tells you the true impact

Step 4: Use a custom model based on your business

Some companies are:

Awareness-limited: Most customers know about you. They just need a push. → Use last-click or time-decay
Consideration-limited: Many people are aware but don’t explore. → Use linear (balanced)
Conversion-limited: Lots of awareness and interest but few buy. → Use first-click (understand awareness)

Know your bottleneck, then weight attribution toward solving it.

Common Attribution Mistakes

Mistake 1: Trusting platform attribution

Facebook tells you “your ads drove 50 conversions.” Google tells you “your ads drove 48 conversions.”

They’re both wrong. They’re both including baseline conversions that would have happened anyway.

(Platform attribution has built-in bias toward that platform.)

Mistake 2: Using last-click for everything

Most businesses default to last-click because it’s easy. But it undervalues awareness.

You cut Facebook (awareness) because it doesn’t get credit. Conversions decline 6 months later because nobody knows about you anymore.

Mistake 3: Comparing channels with different roles

“Our email CAC is $5. Our Facebook CAC is $20. Email is better.”

But maybe:

Email is cheap because you’re emailing people who already know you
Facebook is expensive because you’re building awareness among cold audiences
Fair comparison would be: Facebook → email → conversion (the full funnel)

Mistake 4: Ignoring cross-channel effects

Turning off one channel affects others.

If you turn off Facebook, Google search volume might drop (because people aren’t aware of your brand anymore).

If you turn off email, conversions might drop but so does email unsubscribe costs.

Single-channel attribution misses these ecosystem effects.

Building a Custom Attribution Model for Your Business

We typically recommend:

For B2B (sales-driven):

First-click: 20% (awareness matters)
Middle: 30% (consideration matters)
Last-click: 50% (closing matters most in sales)
Bias: Toward last-click because sales cycles are long and closing requires effort

For E-commerce (self-serve):

First-click: 30% (awareness matters)
Middle: 20% (less consideration; mostly impulse buying)
Last-click: 50% (retargeting and conversion critical)
Bias: Toward last-click because conversions happen quickly

For SaaS (mixed):

First-click: 25% (awareness matters for brand)
Middle: 35% (consideration and evaluation important)
Last-click: 40% (conversion important but not as much as B2B sales)
Bias: Toward middle (balanced across awareness, consideration, conversion)

For Product-Led Growth (self-serve, no sales):

First-click: 40% (awareness critical)
Middle: 30% (engagement matters)
Last-click: 30% (product does the closing, not ads)
Bias: Toward awareness because there’s no sales team to close

The Pragmatic Approach

Most businesses overthink attribution. Here’s what we actually recommend:

1. Calculate last-click as your baseline (it’s simple, everyone understands it)

2. Identify your “hero channels” (which channels show ROI even under last-click model?)

3. Double-check with incrementality testing (turn off each hero channel for a week and measure impact)

4. Allocate budget based on tested impact, not model

The model helps you think through the problem. Testing gives you the answer.

The Takeaway

Attribution models are useful for thinking through channel strategy, but none are perfect.

Don’t pick one model and trust it religiously. Use multiple models to triangulate. Run incrementality tests to validate.

Understand that each channel has a role (awareness, consideration, conversion). Weight attribution accordingly.

We help you build a custom attribution model based on your business and validate it through testing.

The goal isn’t to find the “perfect” attribution model. It’s to make smarter decisions about where to spend marketing budget.