After Google's August Bidding Update, Remove the Targets You Already Beat

Aaron Young of Define Digital Academy analysed account data after Google Ads' August 17 bidding update, and one conclusion runs against what many advertisers are doing: if a campaign is beating its target ROAS, remove the target instead of leaving it in place.

1) The target is now what Google aims for

According to the video, especially for budget-limited campaigns using target ROAS or target CPA, the stated number is now what Google tries to hit. The old trick of capping budget and setting a high target for low volume but fat margins no longer works. Across 502 campaigns from 78 accounts over 90 days, target ROAS campaigns ran at 159% of target before the update, 98% in the first four weeks after, and 89% in weeks five to eight. Readers of why target ROAS is an aggressiveness dial, not a promise will recognise the dial, now far more sensitive.

2) Volatility follows conversion volume per campaign

Below 10 conversions a week, 80% of campaigns swung more than 20% in ROAS and 35% had weeks with zero conversions. At 30 to 50 a week, zero-conversion weeks disappeared. An account with about 80 conversions a month should run no more than four campaigns, ideally two or three.

3) The order of operations

Remove targets you are already beating. Reintroduce one only after at least 10 conversions a week and four to six flat weeks. Consolidate instead of splitting. Do not change bidding after one or two weeks, because Google works in roughly 30-day cycles.

Minh adds a condition from B2B work at QH Distribution: weekly lead volume is far lower than in ecommerce, so splitting campaigns by product or market starves the algorithm. Minh consolidates first and picks a conversion close to revenue, as in optimising Google Ads for lead quality, before setting any target.

The takeaway: after August 17, a target is no longer a ceiling to beat but a point Google pulls you back to.