Google expands use of machine learning for advertisers by launching Responsive search ads, Maximize lift for YouTube, and Local and Smart shopping campaigns
The ways people get things done are constantly changing, from finding the closest coffee shop to organizing family photos.
Context & Ripple Effects
Earlier in 2018, Google had already pointed its machine-learning stack at the supply side of ads with AdSense Auto ads, which automated placement and monetization choices for publishers. This announcement carries the same logic to the demand side: Responsive search ads let Google assemble and test headline-and-description combinations itself, while Maximize lift for YouTube, Local shopping, and Smart shopping campaigns hand bidding and targeting decisions to Google's models.
That direction only intensified afterward — Google extended visual automation with Discovery ads across YouTube, Gmail, and Discover in 2019, then collapsed format-by-format automation into fully automated Performance Max campaigns in 2020, and by 2023 was applying generative AI to Search ad relevance and campaign creation. The 2018 launch is where that arc starts on the advertiser side.
First-order effects
- Advertisers get fewer manual levers: Responsive search ads shift creative control from fixed copy to asset inputs Google mixes and ranks, while Smart and Local shopping campaigns move retail bidding and inventory signals under Google's optimization.
- Search marketers and agencies managing granular keyword-and-ad-group structures face a workflow change immediately, as Google positions its own models as the default optimizer across Search, YouTube, and Shopping.
Second-order effects
- Each automated format becomes a template Google can replicate across surfaces — the pattern that led from these single-format products to whole-campaign automation in Performance Max, further concentrating budget allocation inside Google's systems rather than advertiser dashboards.
- Rival ad platforms face pressure to match the automation pitch, since advertisers comparing platforms increasingly weigh whose machine learning spends the budget better rather than who offers more manual controls.
Third-order effects
- If the pattern holds, the advertiser's role narrows to supplying assets, goals, and budget while Google's models decide matching, bidding, and creative assembly — the endpoint visible in today's conversational campaign building and AI-powered ad formats, with measurement and accountability becoming the main contested ground between platforms and buyers.
The trend: Google has spent the years since 2018 systematically replacing manual ad-buying controls with machine-learned automation, moving from per-format tools like these to fully automated and now generatively created campaigns.