Google Ads Target CPA budget limited: what changed on 17 August 2026
Since 17 August 2026, a "Limited by budget" Target CPA or Target ROAS campaign aims closer to the target you set, even if it had been beating that target.
"Google won't automatically adjust your bidding targets or budgets."
Google Ads Help, "Changes to target based bid strategies"
Google's example is a $10 Target CPA campaign that had been achieving $5. It "will start delivering closer to the $10 USD target you originally set". Your settings did not change, and Google's page lists no opt-out. Check your change history; the page does not say whether this change shows there.
Our hypothesis: some campaigns quietly beating their targets were carrying stale targets. Test each one against your Recommendations forecast, and fix your number where it was. That is core AI performance marketing work for capped accounts.
Why can CPA rise when nobody changed a setting?
The target you set became the number the system aims for. Google gives no CPA figures, so the mechanism is our reading.
Google describes "campaigns currently performing more efficiently than their set targets". Slack or real efficiency depends on whether you could buy volume at the target; your Recommendations forecast is the test. Google does not say whether spend holds at your cap. If it does, your conversions move; if not, your spend moves (our reading; Google says Maximize strategies "optimize to spend your full budget"). Check both.
Your finance team sees cost per lead up by more than half and asks who changed the bids. Google's page does not say whether this change appears in your change history, so check it.
Which campaign types are affected?
Search, Shopping, Performance Max, Demand Gen and Travel campaigns changed on 17 August 2026, so check yours first.
| Campaign type | Affected by the 17 August change | What to check |
|---|---|---|
| Search | Yes | Your actual vs average target CPA, both windows |
| Shopping | Yes | Your actual vs average target ROAS; product groups that lost volume |
| Performance Max | Yes | Your CPA or ROAS gap since mid-August; compare channel mix too |
| Demand Gen | Yes | Your CPA gap and channel mix; was it budget-limited at all |
| Travel | Yes | Your ROAS gap by property or route |
| Display, Hotel | Already behaved this way | Nothing new |
| App, Video reach, Video view | No, "previous bidding behavior" continues | Exclude from your audit |
It applies in Google Ads, Search Ads 360, Display & Video 360, Google Ads Editor and the Google Ads API, so search advertising through SA360 is inside it. Google defines "Limited by budget" as a status where "your average daily budget is lower than the recommended amount". A campaign under its cap every day, or Maximize conversions with no target, is outside the change. One limited some days is affected in proportion, hence the share in the estimator.
How to audit Google Ads Target CPA budget limited campaigns
Compare your 30 days before 17 August 2026 with your 30 days after, and sort your campaigns by the gap between actual CPA and average target CPA.
| Step | Where to look | What a gap means |
|---|---|---|
| 1. List the exposed campaigns | Your Campaigns page, Status column "Limited by budget" | Only your Target CPA or ROAS ones are inside the change |
| 2. Measure the drift | Your bid strategy report: 18 July to 16 August vs 18 August to 16 September 2026 | Actual CPA up toward target, cost flat: the change |
| 3. Rule out your own edits | Your change history, same windows | Any target, budget or conversion edit you made near 17 August confounds the read |
| 4. Check the status | Your bid strategy status | "Learning" means wait; "Limited" means the cap still binds |
| 5. Count lost conversions | Your campaign report, same windows | Fewer conversions at equal spend is what your stale target cost you |
| 6. Decide per campaign | Your Recommendations forecast | Pick one of five options and write down your reason |
The bid strategy report shows "Average Target CPA", "Actual CPA" and "Conversions". Compare against the average target, which Google's Target CPA page defines as "the traffic-weighted average CPA that your bid strategy optimized for".
Google says to "Wait until all your conversions have been reported" before evaluating, and to measure "the last 30 days, including at least 30 conversions"; below that, drift and noise look the same. A comparison cannot prove the cause. Start with your most capped campaigns. An unlimited Target CPA campaign in your account, which the change did not touch, is the cheapest control: if it drifted too, the cause is not this change.
Formula: gap = your target CPA minus your pre-change CPA. On budget-limited days the strategy now aims closer to the target, so expected CPA runs from pre-change CPA + gap × limited share × 0.5 (partial convergence) to pre-change CPA + gap × limited share (full convergence). At a fixed budget, if spend holds, conversions scale with pre-change CPA ÷ new CPA. Defaults are made-up round numbers.
Expected CPA 76 to 92 against your target of 100, a rise of 27% to 53% on your pre-change 60. If your spend holds, expect 21% to 35% fewer conversions; if it falls instead, compare spend too. Illustrative heuristic; confirm in the bid strategy report.
Static version (target 100, pre-change CPA 60)
| Share of days limited by budget | Expected CPA band | Conversions if spend holds |
|---|---|---|
| 25% | 65 to 70 | -8% to -14% |
| 50% | 70 to 80 | -14% to -25% |
| 75% | 75 to 90 | -20% to -33% |
| 100% | 80 to 100 | -25% to -40% |
Should you lower the target, switch strategy or raise the budget?
Lower the target only to a number you can defend, switch strategy only if volume beats a stable CPA, and raise the budget only where the forecast supports it.
Each answer narrows the choice. Nothing is suggested until all four are set.
| # | Google's option | Use it when | Then |
|---|---|---|---|
| 1 | Keep your current target as it is | A lead is worth your target and your volume rose | Report your volume |
| 2 | Adjust your target in line with recent performance | Your old, lower CPA was the plan | Re-check after one conversion cycle |
| 3 | Adjust your target to a custom number | Neither of your numbers is honest; Google's example is $7 | Build it from cost per qualified lead or bound policy |
| 4 | Change your bid strategy | Your cap is the real limit and volume beats a tidy CPA | Maximize strategies "optimize to spend your full budget" |
| 5 | Increase your budget | Your forecast shows conversions at your target | "wait 1 to 2 conversion cycles" |
Google adds that on a Maximize strategy "your actual CPA or ROAS will fluctuate as you adjust your budget". Option 2 is the closest Google offers to a rollback. We object to it for the reason below, and in our view the custom target in option 3 is usually the honest one.
If you lower every target to last quarter's actual CPA, you feel you are restoring the old behaviour. You are not: the old behaviour beat a loose target by accident, and your new setting hard-codes the accident as the goal. If demand softens, a campaign pinned to a low target shrinks while your budget sits unspent.
What should finance see before year-end?
Tell your finance team that your cost per lead may have moved on 17 August 2026, likely because Google changed how budget-limited targets behave, and show them both windows.
| Row | What goes in it |
|---|---|
| Pre-change window | Your actual and average target CPA, conversions, cost |
| Post-change window | The same columns, after conversions finish reporting |
| Decision per campaign | Keep, match, custom, switch or fund, with your reason |
| New target | Your business rationale; finance owns the claim |
Re-baseline Q4 and full-year targets from the first clean post-change window. Low lead volume makes drift noisy, so insurance marketing teams should read the windows with care, and so should you if your capped account sits in Singapore, Malaysia, Australia, the US or Canada. Our insurance paid search guide covers the conversion set-up that makes a target meaningful.
Finance asking why a capped account's CPA rose in August? Leapbuzz builds the before-and-after audit and the re-based targets: ask for the audit.
The August step-change shows a target drifted from the business; it says nothing for or against bidding automation. The board-grade ROI proof format fits the pack. If the rise still cannot be squared with the pipeline, our incrementality testing guide covers geo and holdout designs. Marketing guidance, not financial or legal advice.
Do this today
Compare actual CPA for 18 July to 16 August with 18 August to 16 September 2026 on every "Limited by budget" Target CPA and Target ROAS campaign.
- Do this first: rank the list by the gap in actual CPA between the two windows.
- Any campaign of yours whose actual CPA rose more than 15% toward target with cost flat gets a decision this week.
- The 15% line is our threshold, not Google's; below it, lag and seasonality explain as much as the change.
- Write your new target and its reason into the finance pack the same day.
The target you leave in place is the price you will pay. Re-check every Google Ads Target CPA budget limited campaign against your Google Ads conversion values once a quarter, starting with this one.
