Google Ads has published five practical recommendations for working with its two agentic tools, Ads Advisor and Analytics Advisor, framing them as conversational assistants rather than dashboards.
The distinction matters. Both tools accept natural-language prompts, remember prior exchanges within a session and refine their output through follow-up questions. The pitch is that advertisers can move between data and decisions without writing a line of code.
What each tool does
Analytics Advisor is positioned as a proactive data analyst. It surfaces trends, runs ad-hoc computations and handles funnel analyses and channel investigations on request. Google suggests starting simple, asking something like "how many new users did we get last week?" before asking the tool to run a deeper analysis. The idea is to build familiarity before delegating complexity.
Ads Advisor sits closer to execution. It diagnoses campaign delivery problems, including policy issues that may be limiting reach, and generates creative assets: keyword ideas, headlines and descriptions. It is a diagnostic and drafting tool in one.
How Google wants advertisers to use them
The recommendations follow a consistent thread. Start with simple queries. Use follow-up questions to go deeper. Validate suggestions before applying them. Use the thumbs-up and thumbs-down feedback controls so the models adapt to each business over time.
That last point is significant. These tools improve with use, which means advertisers who engage with them consistently will get different outputs than those who treat them as one-off queries. The feedback loop is part of the product.
Google is candid about the limitations. The announcement notes that content from both advisors is generated by AI and that generative AI remains experimental. That caveat sits in tension with the collaborative framing, but it is the honest position.
The broader shift
Related reading
- Microsoft warns AI agents risk becoming "double agents" as it unveils security controls at RSAC
- Meta hires founders of AI startup Dreamer in latest push to build autonomous agent technology
- Nvidia says the open versus proprietary AI debate is the wrong argument
What Google is describing is a change in how advertising decisions get made, moving from a workflow where humans pull reports and draw conclusions to one where an AI surfaces the insight and proposes the action. The human role becomes review and judgment rather than analysis.
Whether that is a compression of skilled work or a genuine productivity gain depends on how well the tools perform in practice. Google's recommendation to treat the advisors as collaborators and review everything before acting is sensible guidance for tools that are, by the company's own admission, still experimental.
The recap
- Google Ads publishes 5 best practices for agentic advisor collaboration.
- Analytics Advisor surfaces hidden trends and builds funnel analyses.
- Use thumbs up or thumbs down feedback to improve recommendations.
