Increase product activation

Auto Ads used ML to place ads and find inventory site owners missed, but most publishers never turned it on. The product was one snippet, yet it felt opaque if you weren't technical. Manual placements were familiar; the ML piece sounded like a black box. Google's bet on automated monetisation only paid off if people actually finished setup, and competitors were happy to pick up anyone who stalled.

We used email to close the gap: segmented sends so each publisher saw copy that matched where they were stuck: first what Auto Ads actually did, then how to paste the tag, then what to check in reporting. Side-by-side examples showed ML placements next to manual ones; follow-up messages walked through the code with screenshots. Publisher stories with real revenue lifts went out as proof, and deeper mails covered mobile anchor ads, AMP, and cross-device setups for people ready for more.

Completion rates on onboarding went up: people finished the flow once the emails demystified the ML piece, and support tickets dropped. More publishers flipped Auto Ads on, so Google's models could surface inventory that manual setups skipped, page by page, at scale. AdSense kept its lead; publishers spent less time babysitting placements and more time on their sites.