"Instead of using prebuilt filters to discern various types of queries, users can enter prompts, such as: "Show question-like queries." "Filter queries longer than [xx] characters." "Show informational queries." The feature initially responds to the prompts with regular expressions. For example, for the prompt, "Show queries with informational intent," it created the following regex: Queries matching regex: (what|how|why|guide|tutorial|explain|definition|example|best way to|steps to|tips for|reasons why|benefits of|difference between) I could then click "Apply" to see the filtered results."
"Similarly, I prompted "Show product research queries." The feature responded: Queries matching regex:(best|review|compare|vs|price|buy|alternatives|guide|recommendations|features|specs|cost|discount|deal|sale|where to buy|how to choose|top rated|pros and cons) Again, I could "Apply" or "Dismiss" that regex. Brand search Queries matching regex: (brandname|brand name) The AI analysis performed surprisingly well on brand-name searches. I prompted, "Show branded queries." The regex responses included my brand name and a one- and two-word pattern:"
Google Search Console's AI analysis in the Performance > Search results tab enables users to enter natural-language prompts to analyze organic search performance. The tool converts prompts into regular expressions that filter queries by intent, length, format, or branding. Generated regex patterns can be applied or dismissed to refine results. The feature supports product-research and question-like query identification, assembles traffic-change reports comparing date ranges, and provides country-specific metrics such as clicks, average CTR, and average position to evaluate search visibility across dimensions.
Read at Practical Ecommerce
Unable to calculate read time
Collection
[
|
...
]