If your Google Ads or GA4 reports show different numbers than last month and you did not touch anything, it is not a bug: Google changed the rules. As of mid-July 2026, the first click, linear, time decay and position-based attribution models are no longer available. Two remain: data-driven attribution and last click.
Practical translation: the way Google decides which campaign gets credit for each sale just changed, and your numbers will change with it — even if your operation stays identical.
What changed, exactly, and when
- From mid-July 2026: new conversions can no longer use first click, linear, time decay or position-based. Only data-driven or last click.
- By September 2026: the four models disappear entirely, including for conversions that already used them. Anything still configured with them migrates automatically to data-driven.
- What remains in GA4: data-driven (the default), paid and organic last click, and Google paid channels last click.
Watch the automatic migration detail: if you do nothing, Google decides for you. And data-driven is a black box — a Google AI model that distributes credit across touchpoints following patterns you cannot audit.
My honest read: good for Google, awkward for SMBs
The official argument is reasonable: fixed-rule models — all credit to the first click, or equal parts to everything — are arbitrary, and a model trained on real data attributes better. On large accounts, that is true.
The problem is the fine print for smaller companies. Data-driven needs conversion volume to work well. A B2B SMB generating 20 or 30 leads a month does not give the model enough signal, and the result can be an erratic credit distribution that shifts month to month. And since the model is not auditable, you cannot explain to management why the number moved.
What I see across B2B clients makes it worse: 3-to-9-month sales cycles, few high-value conversions, and decisions that happen over WhatsApp, email and meetings Google never sees. Google attribution was always a partial picture; now it is a partial picture you cannot configure.
What to review in your account this week
- Audit your conversions. In Google Ads: Goals → Conversions → attribution model column. Identify which ones still use first click, linear, time decay or position-based. Those will migrate to data-driven on their own.
- Export the model comparison NOW. In GA4, the model comparison report still shows how credit shifts between last click and data-driven. Save that export: it is your baseline for explaining the jumps ahead.
- Document your current numbers. Cost per lead, conversions per campaign, ROAS for the last 6 months. When the model changes, you will need to tell a real performance change from a measurement change.
- Choose deliberately: data-driven or last click. With low monthly conversion volume, last click is less elegant but more stable and explainable. With volume, keep data-driven and monitor the first weeks.
- Strengthen your own measurement. A source question on your forms, source tracking in the CRM, imported offline conversions. The less you depend on Google’s black box to understand where clients come from, the better.
The underlying point
Every time Google simplifies measurement, whoever owns their own measurement wins. Perfect attribution does not exist — least of all in B2B — but an SMB that records in its CRM where every closed deal came from does not depend on whichever model is fashionable this year. If this change forces you to fix that, it will be the most profitable thing to come out of July.
Not sure which model fits your account? Review it calmly before September. And if you want a second opinion on your conversion setup, reach out — it is a one-hour review, not a project.
