Google Ads Performance — CPA & Conversion Rate, YoY

CPA / Conversion Rate correlation — what's actually driving CPA

For each market, this compares how strongly weekly CPA moves with weekly conversion rate (CVR) versus how strongly it moves with weekly cost-per-click (CPC) across the 33 weeks of 2026. A correlation close to −1 between CPA and CVR means CPA rises almost entirely because conversion efficiency falls, not because clicks got more expensive.

• Australia: r=-0.898 (CPA–CVR) vs r=0.58 (CPA–CPC) — CPA swings are driven overwhelmingly by conversion rate, not the other factor.
• United States: r=-0.927 (CPA–CVR) vs r=0.265 (CPA–CPC) — CPA swings are driven overwhelmingly by conversion rate, not the other factor.
• United Kingdom: r=-0.767 (CPA–CVR) vs r=0.685 (CPA–CPC) — CPA swings are driven overwhelmingly by conversion rate, not the other factor.
• Canada: r=-0.87 (CPA–CVR) vs r=0.326 (CPA–CPC) — CPA swings are driven overwhelmingly by conversion rate, not the other factor.
• New Zealand: r=-0.617 (CPA–CVR) vs r=-0.279 (CPA–CPC) — CPA swings are driven overwhelmingly by conversion rate, not the other factor.

Reading this across markets: in AU, US and CA, CPA is overwhelmingly a conversion-efficiency story (r between −0.87 and −0.93 vs CVR, weak r vs CPC) — auction cost has little to do with it. UK is the exception — both CVR (r=−0.77) and CPC (r=0.69) move CPA meaningfully, so click cost is playing a real role there too, not just funnel efficiency. This supports treating the CPA problem primarily as a conversion-signal issue (tracking, checkout, funnel) rather than a bidding-competition issue — with UK needing a look at auction dynamics as well.