Both lift the $2.00 bid cap. Both optimise toward conversions. The difference is that Target CPA adds a constraint — and on a grant account, where the budget is donated and unspent budget is simply lost, that constraint is usually the wrong thing to add.
The difference in one line
Maximize Conversions asks: get me as many conversions as this budget allows. Target CPA asks: get me as many conversions as you can at roughly this cost each, and decline the ones that cost more.
On a commercial account paying real money, that second constraint is essential — an acquisition costing more than the customer is worth is a loss. On an Ad Grant, the money is donated and expires monthly. A conversion that costs $18 instead of $9 costs the organisation nothing extra.
Default to Maximize Conversions. Most grant accounts should never use Target CPA. Setting a target on free budget is choosing to leave some of it unspent in exchange for an efficiency figure nobody is paying for.
When Target CPA does earn its place
Three situations, all of which are signs of a mature account.
- Volume is stable and high. Roughly 30+ conversions a month per campaign. Below that, Target CPA has too little data and throttles serving badly.
- Conversion quality varies by cost. If your cheap conversions are newsletter sign-ups and your expensive ones are major-gift enquiries, a target can steer toward one band. Though separating them into different campaigns usually works better.
- The grant maxes out. If you are genuinely spending near $10,000 a month, budget is now scarce and efficiency starts to matter. At that point you have the problem Target CPA was designed for.
That last one is rare. Most accounts are trying to spend more, not better.
The mistake that stops an account dead
Setting a Target CPA lower than the account can actually achieve. Google will not bid above what the target implies, so if your real cost per conversion is $22 and you set a $6 target, the campaign stops serving almost entirely. Impressions collapse, spend goes to nearly zero, and it looks like something broke.
If you do set one, take the actual cost per conversion from the last 30 days and set the target at that number, not below it. Tighten later in 10–15% steps, allowing two weeks between changes. Never set an aspirational target.
Moving between them
The sequence that works for an account earning its way up:
- Manual bidding while conversion tracking is installed and verified.
- Maximize Conversions once two to three weeks of clean data exist. This is where most accounts should stop.
- Target CPA only if volume becomes stable and budget becomes scarce.
- Target ROAS only if conversions carry genuine monetary values passed back from the donation platform. If you are inventing an average gift value to make the maths work, do not use it.
Each change restarts a learning period. Do not move between strategies more often than monthly, and never mid-campaign on a time-limited appeal.
Reading the result honestly
Switching from Maximize Conversions to Target CPA will usually show a lower cost per conversion. That is not evidence it worked — it is arithmetic, because you told Google to decline expensive conversions. The question is whether total conversions held.
| After the switch | Verdict |
|---|---|
| CPA down, conversions flat | Genuine efficiency gain. Keep it. |
| CPA down, conversions down | You bought a number. Revert. |
| CPA down, spend collapsed | Target set too low. Raise it or revert. |
On a grant account the middle row is the common one, and it is why the default recommendation holds.
The short version
Use Maximize Conversions. Revisit the question only if you are spending close to the full grant, which is a good problem and a rare one. And if you do set a target, set it at reality rather than at ambition.