Google confirmed in a mid-July 2026 update that four of its attribution models are gone. First-click, linear, time-decay, and position-based — all removed. If your campaigns are still running on any of these, you have until September before they’re force-migrated to data-driven attribution.
Two dates matter. Starting mid-July 2026, these four models stopped being available for any new conversion action. By September 2026, they disappear entirely — including for conversion actions that were still using them.
Once that migration finishes, exactly two attribution models remain in Google Ads: data-driven attribution and last-click.
Why Google Killed These Models
Google’s reasoning is straightforward: the old models were rules-based. They assigned credit based on fixed logic rather than actual data. First-click gives 100% credit to the first touchpoint. Linear splits it evenly. Time-decay gives more to recent touches. Position-based gives 40% to first, 40% to last, 20% to middle.
None of these reflect how people actually convert. Data-driven attribution uses machine learning to analyze what actually contributed to conversions across your specific account. It looks at thousands of data points to determine which touchpoints matter most.
The result: credit is assigned based on evidence, not assumptions.
The Bigger Platform Shift
Google isn’t the only platform narrowing how credit gets assigned in 2026:
- January 2026: Meta permanently removed the 7-day and 28-day view windows from its Ads Insights API. Industry analysis put the resulting conversion drop at 15 to 30 percent for accounts that leaned on those longer windows.
- March 2026: Meta narrowed what counts as a click-through conversion. Likes, shares, and saves no longer trigger the 7-day click attribution window — only link clicks do.
- July 2026: Google removes four attribution models, leaving only data-driven and last-click.
Stack these changes together and the shape of the year becomes clear: the platforms are simplifying measurement on their own schedule. Reported numbers move without any change in real performance, and the advertiser has no visibility into the model doing the deciding.
What This Means for Your Campaigns
If you were using first-click, linear, time-decay, or position-based attribution, here’s what happens:
- Your reported conversions will change. Different models assign credit differently. When you switch from linear to data-driven, some campaigns will look better, others worse — even though nothing changed in your actual performance.
- Campaign budgets may shift. Google’s Smart Bidding uses attribution models to decide how to distribute spend. A model change means the algorithm sees different data and distributes budget differently.
- Historical comparisons break. You can’t compare pre-September performance (old model) to post-September performance (data-driven) and call it a real trend. It’s a measurement artifact.
What To Do Before September
- Audit your current attribution setup. Go to Tools & Settings > Conversions > Attribution Model. Check every conversion action. If any are set to first-click, linear, time-decay, or position-based, note them down.
- Export your current performance data. Pull reports for the last 90 days using your current model. This is your baseline. After September, you’ll need this to understand what actually changed vs. what’s just a model shift.
- Switch to data-driven manually. Don’t wait for Google to force-migrate you. Switch now so you control the timeline and can monitor the impact gradually rather than getting a sudden shock in September.
- Rebuild your reporting baselines. Create a new reporting view that starts from the day you switched. Compare this forward, not backward.
- Communicate with stakeholders. If your team or clients are used to seeing specific conversion numbers, prepare them for the shift. Explain that real performance hasn’t changed — just how it’s measured.
The Honest Truth About Attribution
Attribution has always been imperfect. Every model is an approximation. Data-driven is better than the old rules-based models, but it’s still not perfect. It can’t account for offline touchpoints, word-of-mouth, or the ad someone saw on a billboard that planted a seed.
The best approach isn’t to obsess over which attribution model gives you the prettiest numbers. It’s to use attribution as one signal among many — alongside incrementality testing, marketing mix modeling, and good old-fashioned business judgment.
Google killed four models. But the real question isn’t which model you’re using. It’s whether you’re making decisions based on a single number or on the full picture.
The shift to data-driven attribution is inevitable. The advertisers who prepare for it now — by exporting data, communicating with stakeholders, and building new baselines — will navigate it smoothly. The ones who get force-migrated in September will spend weeks explaining why their numbers changed.
How We Help Clients Navigate This
At The Virtual Marketing Era, we’re already migrating all client accounts to data-driven attribution and rebuilding reporting baselines. If your campaigns are affected by this change and you want to make sure the transition is seamless, book a free strategy call. We’ll audit your attribution setup and build a migration plan that protects your performance data.