Key takeaways
- Google’s target bidding changes make Target CPA and Target ROAS behave more like literal performance targets, not safety nets.
- The biggest risk is setting targets that do not match campaign objectives, margins or lead quality.
- Conversion tracking faults make Smart Bidding more expensive because Google optimises towards the signals marked as primary.
- UK advertisers should split campaigns where brand, non brand, product margin or customer value economics differ.
- Move CPA and ROAS targets gradually, then judge results after a full conversion cycle rather than reacting daily.
Google’s target bidding changes are not a reason to panic. They are a reason to stop treating Target CPA and Target ROAS as magic efficiency guards. The money moves when the target you set becomes the level Google actively works towards, rather than a ceiling it is happy to beat.
For UK advertisers, the danger is overreaction. Teams will see CPA rise towards a target, assume Smart Bidding has broken, then slash budgets or reset strategies before the data has matured. That is the wrong move. The right one is to audit objectives, tracking and structure before touching a single target. If your account already leans heavily on automation, our take on Google Ads automation signals is a better starting point than another round of bid panic.
The short version: these changes reward advertisers who know their numbers. They punish accounts using tidy looking targets that bear no relationship to margin, lead quality or conversion lag.
What target bidding changes actually changed
Target CPA and Target ROAS are being framed more literally. Previously, a campaign with a £50 Target CPA that could deliver leads at £35 often carried on doing so. The target acted like an efficiency safeguard: if Google found cheaper eligible conversions, it took them.
Under the newer interpretation, the target is closer to the outcome Google is trying to average around. A £50 Target CPA gives the system permission to bid into auctions that bring performance towards that number. Some conversions land below it, some above it, but the overall aim is no longer to beat the target wherever it can.
That sounds dramatic, but PPC teams have managed this before. Early Target CPA worked in much the same way. What has changed is the environment around it: broader match types, Performance Max, Demand Gen, AI assisted campaign creation and far less manual query control. The mechanism is familiar. The blast radius is larger.
Why this matters for advertisers
The biggest commercial risk is not that Google has changed a label. It is that advertisers set targets that do not match the actual job of the campaign. If your real objective is maximum qualified lead volume from a fixed monthly budget, Target CPA is the wrong starting point. It tells Google to prioritise efficiency around a number, even when profitable volume is available above or below that line.
Here is the mechanism. A campaign previously delivering leads at £38 against a £50 Target CPA had a £12 efficiency cushion. Finance liked it, sales liked it, and the dashboard looked strong. If Google now works more directly towards £50, CPCs rise because the system enters more auctions where it believes conversions are available at that allowed cost. Volume rises only if those auctions contain genuine incremental demand. If they contain lower quality leads, you have paid more to make the account look compliant with its own target.
Target ROAS has the same problem in reverse. Ecommerce advertisers often set a ROAS target based on blended revenue, then wonder why campaigns stall when the target is too aggressive for non brand demand. Google does not know your warehouse constraints, returns rate, supplier rebates or contribution margin unless you pass better value data back. It optimises against the values it receives, not the values in your finance system.
The tracking problem sits underneath the bidding problem
This is where the UK market has a real weakness. In our audit work we routinely find conversion tracking faults serious enough to distort the numbers a campaign optimises on: duplicated conversions, soft actions marked as primary, and tag firing gaps behind Consent Mode. If you are not confident yours is clean, our guide to what conversion tracking is and why it matters is the place to start.
That matters because Smart Bidding does not optimise for truth. It optimises for the conversion actions you mark as primary, the values you import and the lag you allow. Feed it spam form fills, duplicated phone calls or soft micro conversions, and it will buy more of them at the target you set. These changes make that more expensive, because a bad target becomes more literal.
Structure matters as well. Brand and non brand traffic should not share one efficiency target. Brand converts cheaply because the user already knows you. Non brand carries auction pressure, competitor comparison and weaker intent. When both sit in one campaign, the blended CPA looks acceptable while non brand spend quietly deteriorates. We see the same issue in Performance Max when branded demand props up ROAS, which is why our guidance on Performance Max brand leaks applies directly here.
PPC Geeks’ View
The specific problem advertisers will face is target drift dressed up as stability. The campaign will appear to be doing what you asked: spending, converting and averaging near the target. The business problem is that the target itself is often wrong. It was set months ago, copied from a previous agency, or based on a blended CPA that includes brand, remarketing and poor quality leads.
We see this most often in lead generation accounts running broad match with Smart Bidding and weak offline conversion feedback. Google receives a lead submission as a success, then treats a sales qualified opportunity and a student asking for a brochure as equal. With these changes, that bad signal gets an approved price tag.
Do not respond to this by chasing the interface. Respond by proving the target deserves to exist. If the commercial number is wrong, every bidding strategy becomes a faster way to buy the wrong outcome.
— Lee Sinclair, Head of Operations, PPC Geeks
The immediate takeaway is simple: freeze major target changes until you have checked conversion quality, target maths and campaign segmentation. This is exactly the kind of issue we look for in a free Google Ads audit, especially where automation is spending against weak tracking or inherited campaign structure.
What advertisers should do next
Start with the objective, not the bid strategy. Label every active Search, Shopping and Performance Max campaign as one of three jobs: maximise conversion volume from a fixed budget, protect an efficiency threshold, or acquire a specific type of customer. If the job is volume, test Maximise Conversions or Maximise Conversion Value without a target. If the job is efficiency, Target CPA or Target ROAS belongs in the conversation.
- Calculate the real target this week. For lead generation, work backwards from close rate, average order value and gross margin. If one in five qualified leads becomes a £4,000 customer and your acceptable acquisition cost is £800, your lead target is not complete until you define what counts as qualified. For ecommerce, split targets by margin group rather than using one blended ROAS across the catalogue.
- Audit primary conversions before changing bids. Go to Goals, Conversions, Summary in Google Ads. Remove soft actions from primary status if they do not represent commercial progress. Form starts, page views and unqualified calls belong in secondary reporting unless they are proven predictors of revenue.
- Split campaigns where economics differ. Separate brand from non brand. Separate high margin from low margin products. Separate new customer acquisition when its allowable CPA differs from returning customer sales. Do not ask one campaign to optimise towards two business models.
- Move targets in controlled steps. If actual CPA is consistently below target and conversion volume is stable, reduce the Target CPA by 10% to 20%, then wait one or two full conversion cycles before changing it again. For ROAS, raise the target in the same measured way. Daily tinkering restarts learning before the auction data has settled.
- Use impression signals to diagnose restriction. Pull impression share, lost impression share due to budget and total impressions into your report. If impressions collapse after a target change, Google has decided the auction cannot meet your number at scale. If CPC rises while conversion quality falls, the target is buying more expensive traffic without better outcomes.
- Step down the bidding ladder when delivery dies. If Target CPA stops producing enough conversions, remove the target and run Maximise Conversions to test whether the target is the restriction. If the campaign still lacks data, use Maximise Clicks with strict query and budget controls, then work back towards conversion bidding once volume returns.
For reference, Google’s own documentation says Target CPA bidding aims to get as many conversions as possible at or below the target you set, while Target ROAS bidding uses conversion value signals to work towards your return goal. If you want a second opinion on delivery problems, our notes on Google Ads bid management walk through the same diagnostics.
What this means for your campaigns
The advertisers who lose money here will be the ones who confuse control with comfort. A named target feels controlled, but it only protects profit when the target is commercially correct, the conversion data is clean and the campaign structure lets Google distinguish cheap branded demand from harder incremental growth.
Do not rip out Smart Bidding because the language has shifted. Do not blindly accept the target Google is now working towards either. Treat the update as a forcing function: prove your CPA, prove your ROAS, prove your conversion quality, then adjust targets slowly. If you want support from a specialist Google Ads agency, the first question should be about data quality, not bidding tricks.
Want a no-nonsense view of what to change first? Start with a free Google Ads audit from our team. For the detail behind this, see Google’s FAQ on changes to target-based bid strategies.
Frequently asked questions
What are Google’s target bidding changes?
They change how advertisers should think about Target CPA and Target ROAS. The target is treated more literally as the average performance goal Google works towards, rather than a loose safeguard the system always tries to beat.
Should I stop using Target CPA?
No. Keep Target CPA when efficiency is the campaign constraint and your conversion data is clean. If the objective is maximum volume from a fixed budget, test Maximise Conversions instead.
How often should I change a Target CPA?
Change it only after one or two full conversion cycles. If performance is stable and actual CPA is below target, adjust by 10% to 20% rather than making daily edits.
Why does conversion tracking matter more now?
Smart Bidding buys more of whatever you define as success. If spam leads, duplicate calls or soft actions are primary conversions, Google will optimise towards them at the target you set.
Should brand and non brand use the same Target CPA?
No. Brand and non brand traffic have different economics. Combining them creates a blended CPA that hides whether non brand spend is profitable.
