Google Smart Bidding guidance returns to one-to-two conversion cycles

What changed

Google’s current Smart Bidding guidance describes learning as taking “1-2 typically” in relation to conversion cycles, according to PPC News Feed’s report of the reverted page. This corrects the Week 39 account, which said Google’s guidance described one to two conversion cycles rather than 50 conversions. That interpretation was already inaccurate when published because Google had changed the wording on 22 September.

What changed and when

Search Engine Land reported that Google changed the page on 22 September to say learning could take “up to roughly 50 conversion events or three conversion cycles”. PPC News Feed later reported that the page had returned to “1-2 typically”. The Performance Max note cited previously also no longer appears. The dates describe different states of Google’s documentation, rather than two compatible rules: 22 September relates to the introduction of the roughly 50-event wording, while the 4 October report concerns its subsequent removal.

The earlier figure should therefore not be presented as Google’s current recommendation. Nor does the reverted wording prove that every campaign completes learning within two cycles. The sources document changes to a help page, not observed stabilisation across a representative set of campaigns.

An editorial test for campaign reviews

Editorial suggestion: define the account’s conversion cycle before selecting a review date. Record the typical interval between an ad interaction and the measured conversion, together with the attribution setting and the date of any material campaign change. A campaign with a long decision process should not be judged on the same calendar timetable as one with rapid conversions.

Use a change log to separate adjustments to bids, budgets, targeting and measurement. If several variables move during the same review window, a later performance change cannot confidently be attributed to Smart Bidding learning. The useful comparison is between equivalent periods containing complete conversion cycles, not progress towards a universal conversion count.

How to use the guidance

Check the live Google Ads Help wording when planning a test or explaining learning status. Treat the documentation as a planning reference and combine it with the account’s own lag and change history. Do not use either the removed 50-event formulation or the current typical range as a guarantee of an individual campaign’s outcome.

Why it matters for advertisers

Using a superseded 50-conversion benchmark could delay decisions or create false confidence about campaign readiness. Cycle-based planning better reflects differences in conversion lag, but it still requires account-level evidence and does not guarantee stabilisation.