12 campaigns, 4 countries (US, CA, AU, UK) by 3 device types, same 5 creatives, voice landing. Window June 16 to 17, 2026.
| Country | Spend | Impr | Clicks | CTR% | Regs | $/reg | Reg/click% | Verdict |
|---|---|---|---|---|---|---|---|---|
| Canada | $181 | 3.43M | 755 | 0.0220 | 25 | $7.26 | 3.31 | Top |
| UK | $171 | 2.79M | 500 | 0.0179 | 11 | $15.59 | 2.20 | Solid 2nd |
| US | $183 | 2.84M | 542 | 0.0191 | 7 | $26.17 | 1.29 | Weak |
| Australia | $11 | 36K | 10 | 0.028 | 0 | n/a | 0 | No inventory |
Cost per registration, shorter is better. Canada wins on both volume and intent (clickers convert at 3.31%, more than double US). US drew clicks but they converted at half Canada's rate. AU had almost no inventory under the targeting, so it cannot be judged.
| Device | Spend | Impr | Clicks | CTR% | Regs | $/reg | Reg/click% | Verdict |
|---|---|---|---|---|---|---|---|---|
| Android | $182 | 3.96M | 901 | 0.0228 | 22 | $8.29 | 2.44 | Best scale + cost |
| iOS | $184 | 3.92M | 441 | 0.0113 | 11 | $16.75 | 2.49 | Expensive clicks |
| PC (Win+Mac) | $180 | 1.22M | 465 | 0.0380 | 10 | $18.00 | 2.15 | High CTR, low volume |
Cost per registration, shorter is better. Android delivered both the most volume and the cheapest registration. iOS had the same impressions as Android but half the clicks (lowest CTR), so its clicks are scarce and pricey. PC clicks best per impression but inventory is limited (1.2M vs 3.9M on mobile).
| Site | Impr | Clicks | CTR% | Regs | Note |
|---|---|---|---|---|---|
| Pornhub | 8.47M | 1620 | 0.0191 | 42 | 96% of delivery |
| Youporn | 340K | 111 | 0.0326 | 1 | High CTR, tiny volume |
| Redtube | 242K | 64 | 0.0264 | 0 | High CTR, tiny volume |
| Tube8 | 29K | 7 | 0.0244 | 0 | Negligible |
We did not restrict placements, so TrafficJunky pushed 96 percent of everything to Pornhub (42 of 43 registrations). The smaller tubes show higher CTR (0.026 to 0.033 vs 0.019) but on tiny volume. The signal is weak, yet worth a dedicated budget test rather than letting TJ concentrate all spend on Pornhub.
| Creative | Creator | CTR% | Regs | Reg/click% | Note |
|---|---|---|---|---|---|
| Cantgetenough (anchor) | Miss Alexa | 0.026 | 8 | 1.7 | Best CTR |
| CallMeBaby | Yeyeloba | 0.0198 | 10 | 2.8 | Most regs |
| CallMeNow | Alexis Mucci | 0.0175 | 10 | 3.2 | Best intent |
| RingMeNow | Kazumi | 0.0179 | 9 | 2.8 | Solid |
| CallMeLateNight | Lisa Daniels | 0.0184 | 6 | 1.8 | Weakest of voice |
The anchor (a plain explicit creative) wins on clicks. The new call-CTA creatives win on registration rate, but that is most likely ad-to-landing message match (call ad to call landing), not raw creative strength. Treat as directional.
| Rank | Landing | Clicks | Regs | Reg/click% | Verdict |
|---|---|---|---|---|---|
| 1 | v5 | 529 | 16 | 3.0 | Top |
| 2 | v3 | 622 | 14 | 2.3 | Most clicks |
| 3 | v1 | 491 | 13 | 2.7 | Close 2nd on rate |
Click-to-registration rate, taller is better. Each campaign was assigned one of three landings (v5 / v3 / v1). v5 led on both registration count and click-to-registration rate. v3 pulled the most clicks but converted them at the lowest rate, and v1 sat close behind v5 on rate. The spread is small and the landings ran across different geo and device mixes, so read this as directional.
The single most efficient cell from the test above. We kept it running solo from June 16 to 19, 2026, then stopped it. Final numbers below, same TrafficJunky source as the test for a like-for-like read.
| Creative | Creator | Impr | Clicks | CTR% | Regs (TJ) | Verdict |
|---|---|---|---|---|---|---|
| Cantgetenough (anchor) | Miss Alexa | 3.26M | 942 | 0.029 | 29 | Best CTR + most regs |
| CallYourAiGirl | Savannah Bond | 3.45M | 644 | 0.019 | 13 | Most impressions |
| RingMeNow | Kazumi | 1.25M | 255 | 0.020 | 9 | Solid |
| CallMeBaby | Yeyeloba | 1.91M | 416 | 0.022 | 8 | Mid |
| CallMeNow | Alexis Mucci | 1.00M | 185 | 0.018 | 6 | Lowest |
Miss Alexa (the anchor explicit creative) was the clear winner again: best CTR (0.029%) and the most registrations (29). Savannah pulled the most impressions but at a lower CTR. Per-creative registrations are TrafficJunky; the campaign total is higher because TJ does not attribute every impression to a specific ad.
| Source | Registrations | Cost per registration |
|---|---|---|
| TrafficJunky | 72 | $6.74 |
| Everflow (source of truth) | 58 | $8.36 |
| Metric | 11 campaigns combined | Our CA / Android |
|---|---|---|
| Campaigns | 11 | 1 |
| Impressions | 7,598,614 | 12,189,682 1.6x |
| Clicks | 1,455 | 2,719 1.9x |
| CTR | 0.019% | 0.022% higher |
| Registrations | 30 | 58 1.9x |
| Paid conversions | 0 | 0 |
| Spend | $486.17 | $485.20 |
| Cost per registration | $16.21 | $8.37 ~2x cheaper |
| Geo | US, CA, UK, AU | Canada |
| Devices | iOS, Android, PC | Android |
Same budget (~$485). Green marks where the single kept campaign beats all 11 combined.