Before killing or funding a campaign, check three things in order: is there enough conversion data to trust the trend, is Search Impr. Share Lost (Budget) actually high, and is CTR falling while frequency climbs. Only the second one means add budget. The rest means fix creative or restructure — never just spend more.
- Below the platform's own learning-phase minimum, a rising CPA is noise, not a verdict — wait before deciding anything.
- Check Search Impr. Share Lost (Budget) before adding spend. If it's low, budget was never the actual constraint.
- Falling CTR with climbing frequency is a creative-fatigue signal, not a targeting or budget problem — fix the ad first.
Every account has one of these right now: a campaign whose CPA has climbed for two weeks straight, and someone asking whether to kill it or throw more budget at it. Most teams guess, based on whichever number is loudest that morning. We run three checks in a fixed order before deciding either way — skip one and you'll get a confident answer that's still wrong.
Check 1: Do you actually have enough data to trust the trend?
A CPA graph with eleven conversions in it isn't a trend. It's the algorithm still exploring. Google Ads' own guidance for Smart Bidding is roughly 30-50 conversions per campaign, per month, before bidding stabilizes. Meta's equivalent is around 50 optimization events per ad set inside a rolling 7-day window. Below either threshold, the platform hasn't finished learning — and neither have you. Killing a campaign here is a coin flip dressed up as a decision.
If you're under the threshold, the answer is neither kill nor fund. It's wait.
Check 2: Is budget actually the constraint?
This is the check most teams skip straight past. "CPA is up, so cut budget" and "CPA is up, so add budget" are both guesses unless you've looked at Search Impr. Share Lost (Budget) — the metric sitting right in the Google Ads interface that tells you whether the campaign is actually being rationed by spend.
If that number is high, the campaign wants to spend more than it's allowed to, and a rising CPA alongside it usually means demand outstripped budget, not that the channel stopped working. Fund it. If it's low, the campaign isn't budget-constrained at all — every extra rupee you add will chase the same limited pool of clicks the algorithm was already finding, and CPA won't move.
Check 3: Is this a creative problem wearing a budget costume?
If budget isn't the constraint, look at CTR and frequency together. Falling CTR paired with climbing frequency means the same audience is seeing the same ad too many times — fatigue, not a targeting failure and not a budget shortfall. This is the case most often mistaken for "the channel is dying," when the fix is a new hook and a new set of creative, not a new budget line.
We check this before touching structure, because a creative refresh is cheaper and faster than a rebuild, and it's the more common cause than people expect.
If none of the above explains it, the problem is structural
Enough data, budget isn't binding, creative isn't fatigued — and CPA is still climbing. That's not a spend problem. It's a structural one: overlapping ad groups competing for the same auction, a bidding strategy set against the wrong goal, or targeting that's drifted wider than the offer supports. This is exactly the kind of thing a proper audit surfaces before it becomes a "kill or fund" debate in the first place — which is why we run that check first, not last, in how we work with a new account.
Restructure it, or kill it and rebuild clean. Either way, more budget on a structural problem just buys you the same mistake at a higher price.
Related service: Performance Marketing