A technique for offloading buyer-intent classification of a filtered keyword list to ChatGPT instead of manually tagging each term. Export the keyword-plus-search-volume data as CSV and prompt in two steps: (1) ask ChatGPT to identify the top keyword families by buyer intent present in the list, (2) ask it to output a deduplicated CSV that assigns every keyword to exactly one intent category, using a stated priority order from most-specific to most-generic intent to resolve terms that plausibly fit multiple buckets.
The priority-order instruction is what makes the classification deterministic — without it, ambiguous keywords could be assigned inconsistently or duplicated across categories.
Used as the assembly step for Buyer/Shopper Intent Segregation for PPC Campaigns; the resulting intent column is then filtered in a spreadsheet to pull one buyer-intent group at a time into the campaign build-out. A sibling technique, in the same offload-a-keyword-task-to-ChatGPT family as ChatGPT-Assisted Misspelling Generation.