Attribution Models for Marketing ROI

A practical guide to understanding how credit for conversions is assigned across marketing touchpoints. Choose the right model to accurately measure your campaigns.

What Is Attribution?

Attribution answers the question: "Which marketing touchpoint(s) deserve credit for this conversion?" A customer rarely sees one ad and immediately buys. They might see a display ad, click a search result, receive an email, and then convert. Attribution determines how you split the credit (and the revenue) among those touchpoints.

The Five Common Models

1. Last-Touch Attribution

How it works: 100% of credit goes to the final touchpoint before conversion.

Example: Customer sees a LinkedIn ad → reads a blog post → clicks a Google ad → converts.

Credit: Google Ads gets 100%.

Best for:

• Short sales cycles (< 7 days)
• E-commerce / direct response
• When you want to optimize bottom-of-funnel campaigns

Limitation: Completely ignores awareness and nurture touchpoints that made the final click possible.

In the tracker: Set Attribution_Model = "Last Touch" and Attribution_Weight = 1.00

2. First-Touch Attribution

How it works: 100% of credit goes to the first touchpoint that introduced the customer.

Example: Customer sees a LinkedIn ad → reads a blog post → clicks a Google ad → converts.

Credit: LinkedIn gets 100%.

Best for:

• Measuring awareness channel effectiveness
• Long sales cycles where initial discovery matters
• Brand marketing evaluation

Limitation: Ignores everything that happened after initial contact—including the touchpoint that closed the deal.

In the tracker: Set Attribution_Model = "First Touch" and Attribution_Weight = 1.00

3. Linear Attribution

How it works: Credit is split equally among all touchpoints.

Example: Customer has 4 touchpoints → each gets 25% of the conversion value.

Best for:

• When you genuinely don't know which touchpoints matter most
• Balanced multi-channel strategies
• Teams early in their attribution journey

Limitation: Treats an accidental display impression the same as an intent-driven search click.

In the tracker: Set Attribution_Model = "Linear" and Attribution_Weight = 1/N (where N is total touchpoints). For 3 touches: 0.33.

4. Time-Decay Attribution

How it works: More recent touchpoints get more credit. Earlier ones get less. Typically uses a half-life of 7 days.

Example (7-day half-life):

• Touchpoint 21 days ago → 12.5% credit
• Touchpoint 14 days ago → 25% credit
• Touchpoint 7 days ago → 50% credit
• Final touchpoint → 100% credit (then normalize all to sum to 1)

Best for:

• B2B with long sales cycles
• When recency genuinely indicates influence
• Retargeting-heavy strategies

Formula for weight:

Weight = 2^(-(days_before_conversion / half_life))

In the tracker: Calculate weights manually and enter in Attribution_Weight. Weights for a single conversion should sum to 1.00.

5. Position-Based (U-Shaped) Attribution

How it works: First and last touchpoints each get 40% credit. Middle touchpoints split the remaining 20%.

Example with 4 touchpoints:

• First touch → 40%
• Second touch → 10%
• Third touch → 10%
• Last touch → 40%

Best for:

• Businesses that value both discovery and closing
• Balanced B2B and B2C measurement
• When you have clear funnel stages

In the tracker: Set weights manually: first = 0.40, last = 0.40, middle = 0.20 / (N-2).

Choosing the Right Model

Implementing in the Tracker

The Conversion Tracking tab includes Attribution_Model and Attribution_Weight columns. Here's how to use them:

Single-Touch Models (First/Last)

One row per conversion. Weight = 1.00. Simple.

Multi-Touch Models (Linear/Time-Decay/Position-Based)

Create one row per touchpoint-conversion combination. Same Conversion_ID, different Campaign_ID, with fractional weights that sum to 1.00.

Example: Linear attribution for conversion CVR-025 with 3 touchpoints:

CVR-025, MKT-002, ..., Linear, 0.33
CVR-025, MKT-005, ..., Linear, 0.33
CVR-025, MKT-007, ..., Linear, 0.34

Revenue Attribution Formula

In the ROI Summary tab, use attribution-weighted revenue instead of raw revenue:

=SUMPRODUCT(
  (ConversionTracking!B:B=A2) *    -- Match Campaign_ID
  ConversionTracking!K:K *          -- Deal_Value
  ConversionTracking!M:M            -- Attribution_Weight
)

This gives each campaign its proportional share of revenue.

Common Mistakes

1. Using last-touch for everything — undervalues awareness channels, leading to budget cuts that eventually hurt the funnel

2. Mixing models mid-analysis — pick one model for a given reporting period and stick to it

3. Ignoring assisted conversions — a campaign with zero last-touch conversions might assist hundreds of them

4. Not having enough data — attribution is meaningless with < 50 conversions per channel per month

5. Attributing to impressions — only count intentional interactions (clicks, form fills, email opens) as touchpoints

Moving Beyond Spreadsheet Attribution

This tracker handles straightforward attribution well. When you outgrow it, consider:

• Google Analytics 4 — data-driven attribution using machine learning
• Marketing mix modeling (MMM) — statistical approach using spend and outcome data
• Incrementality testing — controlled experiments to measure true lift

For most businesses doing < $100K/month in ad spend, the spreadsheet approach combined with disciplined data entry provides 80% of the insight at 0% of the cost of enterprise tools.