Since 17th August, budget-constrained Target ROAS and Target CPA campaigns bid closer to the target you’ve entered, not the (much better) results they used to deliver. At Wired Media, CPCs are up around 40% across the accounts we manage. In many of them, dropping the target and moving to Maximise Conversion Value (ecommerce) or Maximise Conversions (lead gen) has settled things down.
So what changed?
Before August 17th, the budget acted as a hidden constraint to bidding algorithms, for example take a generic ecommerce campaign with a 400% Target ROAS and a £100 daily budget; Google was allowed to spend that full budget as long as it targeted at least 400% return, but with limited money it became selective meaning the campaign would often over perform and give you an actual ROAS of 600% or even 700%. Plenty of advertisers were able to set a target well below what they needed, knowing the budget would keep actual returns higher.
Google now optimises more consistently towards the target you enter, even when the campaign is limited by budget so if you say 400% is fine, Google has far less reason to protect the 700% you used to get. It will enter pricier auctions and push performance towards 400% – the promise was supposed to work both ways however and often times campaigns would underperform and hit around 200%, this was supposed to stabilise but in practice tROAS strategies now only work in Google’s favour.
Target CPA works the same way. A £50 target on a campaign that historically delivered £30 leads is now permission to pay up to £50.
CPCs are Skyrocketing – What can we do?
Google doesn’t need more budget to do this. If your average CPC goes from £1 to £1.40, a £100 budget still gets spent in full, but you get roughly 71 clicks instead of your previously attainable 100.
That’s what makes the change really hard to spot. Average spend looks normal and the budget is hit perfectly, but traffic has quietly fallen. We’re seeing CPCs around 40% higher overall and advertisers in the PPC community are reporting increases of 20% to 40% or more, especially in budget-limited campaigns – this is an industry wide problem.
Why raising your ROAS target simply isn’t the fix you think it is.
The obvious response is to raise Target ROAS to the level you actually need to turn a profit – however in our experience this simply creates a different problem; A 600% target doesn’t guarantee a 600% return. It restricts the auctions you can enter, cuts traffic and makes delivery incredibly volatile… and the campaign can STILL finish below the target you set. Cutting a Target CPA aggressively risks a spiral: less traffic, less conversion data and a bid strategy that gets harder to optimise.
What’s working for us: let the budget be the guardrail
Ecommerce: Where budgets are fixed and tracking is reliable, we’re personally moving campaigns to Maximise Conversion Value with no Target ROAS. The instruction becomes simple: spend this budget and get as much revenue as you can. In tested accounts, CPC pressure has eased and traffic has recovered – The guardrail of a target ROAS used to exist both ways but with this new change, it’s become apparent that this now only works in Google’s favour, essentially: removing this guardrail stops your campaigns from telling Google that you are happy for a lower ROAS that it’s actually able to achieve for your business.
Lead generation Same logic, slightly different approach: Maximise Conversions with a fixed budget, while we manage lead quality through structure, targeting, negative keywords, landing pages and better first-party data.
ROAS and CPA still absolutely matter: The big difference is using them to judge performance rather than as an auction-level bidding restriction. Target ROAS and Target CPA still make sense where budgets are flexible and you need to hold a specific efficiency level. It’s just that this should no longer be your default strategy in the wake of August 17th.
Diagnose before you change anything
If performance has dropped, work out why before restructuring everything:
* Has CPC risen, or has traffic fallen?
* Has conversion rate changed?
* Have search terms broadened, or has AI Max been switched on?
* Has Performance Max shifted spend into different channels?
* Is the campaign newly budget-limited?
* Has conversion tracking changed?
* Are leads worse, or just more expensive?
The answers tell you whether the problem is bidding, targeting, tracking or even traffic quality. Changing a target without knowing that isn’t optimisation, it’s guesswork.
Structure matters more than ever before
If Smart Bidding will pay 40% more for a click, then the quality of traffic you let it compete for matters even more. Loose keyword structures, overlapping ad groups, weak negatives and generic landing pages all get more expensive and less profitable for your business. We’re putting more emphasis on tightly themed campaigns, clear keyword intent, strong negatives, relevant ad copy and landing pages that match the search.
The takeaway
Don’t ditch automation and go back to Manual CPC. Do, however, stop assuming your old bidding setup behaves the way it did. If a campaign has been on the same target for two years, test it now: run a controlled comparison against Maximise Conversion Value or Maximise Conversions and judge it on profit, not on how close it gets to a number you entered years ago.
And if you found all of this overwhelming and confusing – then please don’t hesitate to get in contact with me at [email protected] and we can take a look through your account structure together with a free mini account audit.
