Experiment context
Hypothesis. Reducing RSA headline count and tightening pins was expected to increase relevance and CR, especially for first transactions, without materially hurting auction position.
How the test ran. Google Ads and GA4 were checked together because platform conversion signals and real transactions could diverge.
Traffic split. 50/50.
Visual summary — positive campaign changes
The simplified, more tightly pinned RSA produced hypothesis-aligned gains in the retained campaign slices below.
Performance results
| Campaign | Ad group | Impr. | Interactions | Interaction rate | Avg. CPC | Cost | Conversions | CPA | Conv. rate | Impr. (Top) % | Impr. (Abs. Top) % | CTR | Conv. value | Avg. CPC | Session Google Ads campaign | Session Google Ads ad group name | Transactions | First purchasers | CPO | CAC | CR (transactions) | CR (first purchasers) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| google_search_eu_game-3_service9_c1 | service9_service6 | 6,556 | 694 | 10.59% | 1.87$ | 1,298.12$ | 24 | 54.09$ | 3.46% | 65.25% | 52.23% | 10.59% | 1,057.96$ | 1.87$ | google_search_eu_game-3_service9_c1 | service9_service6 | 11 | 5 | 118.01$ | 259.62$ | 1.59% | 0.72% |
| google_search_eu_game-3_service9_c1_test-ads-strength | service9_service6 | 6,496 | 684 | 10.53% | 1.89$ | 1,290.83$ | 33.25 | 38.82$ | 4.86% | 67.90% | 53.17% | 10.53% | 1,987.22$ | 1.89$ | google_search_eu_game-3_service9_c1_test-ads-strength | service9_service6 | 20 | 11 | 64.54$ | 117.35$ | 2.92% | 1.61% |
| Change, % | — | -0.92% | -1.44% | -0.57% | 1.07% | -0.56% | 38.54% | -28.23% | 40.46% | 4.06% | 1.80% | -0.57% | 87.84% | 1.07% | N/A | N/A | 81.82% | 120.00% | -45.31% | -54.80% | 84.48% | 123.22% |
| google_search_us_game-3_service9_c1 | service9_service6 | 6,341 | 422 | 6.66% | 3.01$ | 1,269.35$ | 17 | 74.67$ | 4.03% | 50.93% | 39.78% | 6.66% | 386.21$ | 3.01$ | google_search_us_game-3_service9_c1 | service9_service6 | 18 | 6 | 70.52$ | 211.56$ | 4.27% | 1.42% |
| google_search_us_game-3_service9_c1_test-ads-strength | service9_service6 | 5,298 | 348 | 6.57% | 3$ | 1,045.34$ | 17.33 | 60.31$ | 4.98% | 51.09% | 39.69% | 6.57% | 729.02$ | 3$ | google_search_us_game-3_service9_c1_test-ads-strength | service9_service6 | 11 | 6 | 95.03$ | 174.22$ | 3.16% | 1.72% |
| Change, % | — | -16.45% | -17.54% | -1.35% | -0.33% | -17.65% | 1.94% | -19.23% | 23.57% | 0.31% | -0.23% | -1.35% | 88.76% | -0.33% | N/A | N/A | -38.89% | 0.00% | 34.76% | -17.65% | -25.89% | 21.26% |
| Campaign | Ad group | Impr. | Interactions | Interaction rate | Avg. CPC | Cost | Conversions | CPA | Conv. rate | Impr. (Top) % | Impr. (Abs. Top) % | CTR | Conv. value | Avg. CPC | Session Google Ads campaign | Session Google Ads ad group name | Transactions | First purchasers | CPO | CAC | CR (transactions) | CR (first purchasers) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| google_search_eu_game-3_service7 | service7_service6 | 966 | 100 | 10.35% | 2.61$ | 260.68$ | 1 | 260.68$ | 1.00% | 63.49% | 47.62% | 10.35% | 57.78$ | 2.61$ | google_search_eu_game-3_service7 | service7_service6 | 4 | 0 | 65.17$ | N/A | 4.00% | 0.00% |
| google_search_eu_game-3_service7_test-ads-strength | service7_service6 | 1,524 | 160 | 10.50% | 2.67$ | 426.92$ | 7.50 | 56.92$ | 4.69% | 65.36% | 46.62% | 10.50% | 326.56$ | 2.67$ | google_search_eu_game-3_service7_test-ads-strength | service7_service6 | 4 | 3 | 106.73$ | 142.31$ | 2.50% | 1.88% |
| Change, % | — | 57.76% | 60.00% | 1.45% | 2.30% | 63.77% | 650.00% | -78.16% | 369.00% | 2.95% | -2.10% | 1.45% | 465.18% | 2.30% | N/A | N/A | 0.00% | N/A | 63.77% | N/A | -37.50% | N/A |
Conclusions
The simplified, more tightly pinned RSA showed the clearest hypothesis-aligned improvement in EU game-3 service9: conversion rate, transactions and first purchasers increased while CPA and CAC fell. US game-3 service9 also improved platform conversion rate, CPA, first-purchaser rate and CAC, although total transactions declined. EU game-3 service7 improved platform conversion rate and added first purchasers, but flat total transactions and higher CPO make this result directional rather than conclusive. Retain the simplified RSA selectively and validate these gains over a longer period rather than applying it account-wide.