What to check before you touch anything
- Over what timeframe is this "consistent"? A 30–40% miss over 7 days and the same miss over 60 days are two very different situations — short-term noise isn't a reason to act.
- Break CPA down by ad group, keyword, and device — the gap is often driven by one or two problem segments, not the whole campaign.
- Did the set of conversion actions or their assigned value change recently? Adding a cheap micro-conversion, for example, can dilute the stats around your more valuable, higher-cost conversions.
- Check your change history — does the CPA increase line up with a specific event: a landing page change, some ads getting paused, or a jump in competition (check Auction Insights)?
- Make sure you're comparing apples to apples: is CPA calculated on "Conversions" or "All conv."? If optimization runs on one metric but you're comparing against the other, the gap may be an artifact of the comparison, not a real problem.
- Is the campaign still in "Learning" status? CPA is naturally unstable and can run above target during that period.
- Look at your competitive landscape on your own keywords. Manually check the search results (Auction Insights shows overlap and position, but not what competitors' ads actually say) and compare: have competitor offers changed (pricing, discounts, delivery/guarantee terms), have their ad messaging or headlines gotten sharper, has their landing page experience improved? Rising CPA is often not a sign that your campaign is degrading — it can simply mean competitors have become more compelling in the same auction. In that case, the fix isn't your bids or your settings — it's your offer's competitiveness.
Possible approaches
- If the gap is stable and explainable (the auction got more expensive), one option is to gradually bring target CPA closer to reality rather than leaving an unreachable target in place that keeps choking volume.
- You could temporarily pull back budget share from the priciest segments (geo, device, ad group) without touching the target itself — a more surgical move than changing target CPA across the board.
- Google generally recommends not reacting to swings shorter than 1–2 weeks — give the strategy time to settle after the last meaningful change (7–14 days is the standard rule of thumb).
- If the overspend is concentrated in specific keywords or ad groups, it's usually more effective to work on that segment directly (ad copy, landing page, match to search intent) rather than loosening the target for the whole campaign — the same logic as tackling one weak spot instead of "treating" the overall goal.
- If the gap reflects a genuine, industry-wide rise in acquisition cost (an external factor), some businesses look beyond the campaign entirely and revisit the underlying product economics — average order value, lifetime value — since the problem may not be solvable through campaign settings alone.
- If your competitive check shows rivals have genuinely pulled ahead on offer and messaging, chances are the fix isn't a technical tweak to the campaign — it's a new communication angle: a new value proposition, refreshed ad copy/offers, and a rebuilt landing page that matches the changed competitive landscape. It's slower than adjusting bids, but often the only approach that solves the problem at the root instead of masking it through auction settings.