What this Skill helps with

Negative Keyword Miner turns a search-term export into candidate negative keywords grouped by theme, match type, placement level, evidence, and blocking risk.

Who this Skill is for

It is intended for paid-search teams, agencies, ecommerce marketers, lead-generation teams, and founders reviewing query waste without blocking valuable demand.

What you provide

  • Search terms with campaign, ad group, match type, clicks, cost, and conversions
  • Conversion value when available
  • Products, services, locations, audience, and business exclusions
  • Target CPA or ROAS and the review period
  • Brand, competitor, research, and ambiguous terms that require special treatment
  • Existing negative lists and placement rules

What it produces

  • Candidate negatives grouped by theme
  • Exact, phrase, or broad-negative recommendation
  • Campaign, ad-group, or shared-list placement
  • Cost and conversion evidence for each suggestion
  • Blocking-risk notes and protected terms
  • Review queue separated from clear exclusions

How the Skill works

The workflow classifies query intent and relevance, then compares spend and conversion evidence with the business context. It treats ambiguous, brand, competitor, and low-volume terms cautiously and shows what useful traffic each negative could block.

Installation and first use

  1. Download and add the Skill to Claude.
  2. Attach a recent search-term export and existing negative lists.
  3. Describe what the business sells, serves, and intentionally excludes.
  4. State the target CPA or ROAS and any protected terms.
  5. Review each candidate in the platform before adding it.

Example workflow and expected output

Provide 60 days of search terms for a B2B software campaign. Ask for clear negatives, ambiguous terms requiring review, match-type guidance, placement level, spend affected, and a blocking-risk explanation for each recommendation.

Limits and review guidance

Query intent can be ambiguous, and negative matching can block valuable variants. Low-volume terms may lack enough conversion evidence. Review brand, competitor, geographic, research, and adjacent-product terms manually before implementation.

Frequently asked questions

Does it add negative keywords to the account?

No. It prepares a candidate list for an authorized account operator to review.

Can it recommend negative match types?

Yes. It can suggest exact, phrase, or broader blocking based on the evidence and risk.

Should every non-converting query become a negative?

No. Conversion lag, low sample size, relevance, assisted value, and blocking risk must be considered.