Last-Click Attribution Is Lying to You

Last-click attribution has been the default in Google Ads since the platform launched. It works like this: whichever keyword, ad, or campaign a user clicked immediately before converting gets 100% of the credit. Everything before that final click - every awareness keyword, every research ad, every mid-funnel touchpoint - gets zero.

For simple, single-session purchases, this is a reasonable approximation. For any service business with a multi-touch consideration cycle - law firms, B2B services, high-ticket contractors - it is actively misleading.

What Last-Click Attribution Actually Does to Your Campaigns

Consider a realistic legal prospect journey: they search "can I sue after a car accident," read an article (organic, no credit), search "car accident attorney near me" (click Ad A, no conversion), visit the site and leave, search "best personal injury lawyers [city]" three days later (click Ad B, no conversion), return directly to the site and convert via a branded search (click Ad C, conversion).

Under last-click, Ad C gets 100% of the conversion credit. Ad A and Ad B get zero. The algorithm learns to invest more in branded and high-intent bottom-funnel keywords and less in the awareness and consideration keywords that put that prospect in the funnel in the first place. Over months, your awareness spend atrophies. You become dependent on capturing existing demand instead of creating it.

According to Google's attribution model documentation, data-driven attribution is now the default for accounts with sufficient conversion volume. But for accounts that haven't been audited or were set up years ago, last-click may still be active without anyone noticing.

Attribution Model Options

• Last-click: 100% credit to final touchpoint. Simple, wrong for multi-touch journeys
• First-click: 100% credit to first touchpoint. Overvalues awareness, undervalues closers
• Linear: equal credit to all touchpoints. Better than last-click but treats all touches as equivalent
• Time decay: more credit to recent touchpoints. Reasonable for short consideration cycles
• Position-based: weighted credit to first and last touchpoints, remainder distributed across middle. Best for high-consideration services
• Data-driven: Google's ML model distributes credit based on actual conversion path patterns in your account. Requires sufficient data volume to be reliable

The 40/20/40 Position-Based Model

For service businesses with multi-touch consideration cycles, position-based attribution with a 40/20/40 weighting is the model I use most frequently. The logic:

This is the model documented in the Complex Litigation case study - where switching from last-click to 40/20/40 attribution recovered 28% more profitable keyword signal and shifted Smart Bidding's behavior toward a broader, more accurate view of what was driving conversions.

Implementation Steps

1. In Google Ads, go to Tools and Settings > Conversions and select your primary conversion action
2. Under Attribution Model, switch from "Last click" to "Position-based" (or "Data-driven" if your account has 300+ conversions in the last 30 days)
3. Review the Attribution Reports in Google Ads (Tools > Attribution) to see the before/after impact on keyword value distribution
4. Expect the Conversions column in campaign and keyword view to redistribute - keywords that were showing zero conversions under last-click will now show fractional conversion credit
5. Allow 2-3 weeks for Smart Bidding to incorporate the new attribution signal before evaluating performance changes
6. Check the Model Comparison report to understand the delta between what last-click was reporting and what position-based shows - this tells you where you were over- and underinvesting

What Changes After the Switch

The most common immediate observation: keywords and campaigns that seemed to have low conversion volume under last-click are now showing meaningful fractional credit. This is not the system being more generous - it's revealing the conversions those keywords were genuinely contributing to that last-click was hiding.

• Awareness and consideration keywords get budget allocation they were previously denied
• Smart Bidding adjusts its real-time auction behavior based on the fuller picture of conversion contribution
• Campaign-level ROAS and CPA metrics shift as the conversion credit redistributes - don't panic, let the algorithm stabilize
• The Model Comparison report becomes a useful ongoing tool for budget allocation decisions

Combined with offline conversion tracking and server-side GTM, multi-touch attribution creates the most complete and accurate campaign measurement stack available in Google Ads. Each layer addresses a different dimension of the data quality problem - attribution addresses credit distribution, offline tracking addresses outcome quality, and server-side addresses signal completeness.

Written by Andreus Jedd Sarte. Reference: Google's attribution model documentation and Complex Litigation case study with documented 28% attribution recovery results.