Since Google’s major update on 17 August, one question has come up again and again: is Performance Max quietly moving budget away from the Shopping feed? The short answer is no, at least for most retailers. For a smaller group, however, we can see a much more significant shift, and it has affected their ad spend.
At Bidnamic, we analysed hundreds of Performance Max campaigns to find out how widespread this change in budget allocation is, where the budget went, and what it did to performance. Here is what the data shows.
How we analysed the data
We looked at a large set of Performance Max campaigns and split their spend by network across two periods:
- Before the update: 1 June to 16 August
- After the update: 17 August to 21 October
We excluded feed-only PMax campaigns. These run without assets, so they are structurally close to 100% Shopping and have no channel mix to shift. That left 373 multi-channel campaigns, and every figure in this article is based on those.
Most accounts were not affected
The first thing the data tells us is that the update has not affected every account. Just over half of the accounts we investigated (53%) held their Shopping-feed share within 2 percentage points of its pre-update level, which is within normal fluctuation.
For most advertisers, the channel mix inside Performance Max stayed broadly stable. If your account sits in this group, there is no reason to panic or restructure.
A minority of retailers lost significant Shopping share
The picture changes when you look at the accounts that did move. Around 22% of the accounts in our analysis lost more than 10 percentage points of Shopping-feed share after the update. At the extreme end, 9% lost more than 25 points. In these accounts, the product feed was close to being switched off, even though Shopping is typically the highest-converting surface for ecommerce retailers.
That is a meaningful change for any retailer relying on Performance Max to put products in front of high-intent shoppers.
Where the budget went
In the most affected accounts, Shopping-feed share moved mainly into Search, and a smaller subset also saw Discover rise.
One account shows how far this can go: Discover reached roughly 80% of spend after the update. When the bid strategy was reverted to Maximise Conversions, the mix returned to the Shopping feed within one week.
That detail matters. It suggests the shift is not fixed, and that bid strategy settings can influence where Performance Max chooses to spend.
The performance cost in the hardest-hit accounts
In the most affected accounts, the shift in budget came with a ROAS drop of approximately 13%. Before the update, Discover was performing efficiently on a small share of spend. As its spend scaled roughly threefold, its ROAS roughly halved. A channel that works well at low volume does not necessarily hold that efficiency when Performance Max pushes more budget into it.
The post-update period is still maturing across all accounts, so the ROAS figure is an early estimate, and the gap is likely to narrow as late conversions come in.
