Negative Keyword Manager

When to use this

• A campaign spends on search terms that clearly do not match the product.
• ACoS is high and the search term report is full of junk terms.
• An auto or broad campaign is bringing irrelevant traffic.
• A seller wants a repeatable rule for what to negate, run on a schedule.

The framework. The Search Term Sieve

Exact versus phrase negatives

• Negative exact. Blocks only that precise term. Use it for a specific drain term that you want gone while keeping close variations live.
• Negative phrase. Blocks any search term containing that phrase. Use it for a whole irrelevant theme (a wrong-product word, a wrong-audience word).
• The danger: a broad negative phrase can silently block converting terms that happen to contain the phrase. Before adding a negative phrase, scan the report for any converting term that contains it. If one exists, use negative exact instead.

Step by step

1. Collect inputs. The search term report with spend, clicks, orders, and sales per term, the product, and the target ACoS or cost per acquisition.
2. Compute the drain threshold from the target cost per acquisition.
3. Classify every term into the four buckets.
4. Protect the winners. List every converting term. Before any negative phrase is added, confirm it would not block one of these.
5. Choose negative type per term. Exact for a single drain term, phrase for a whole irrelevant theme that is safe to block.
6. Graduate the winners. Converting search terms found in auto or broad campaigns should be added as exact-match keywords in a harvest campaign, not just left.
7. Set a cadence. Negative keyword work is recurring. recommend a weekly or biweekly pass.
8. Run the quality check, then deliver.

Output format

## Negative Keyword Plan. [campaign]

Drain threshold: spend over [$] with no sale

### Negatives to add
[term] . [negative exact / phrase] . reason . spend wasted [$]
...

### Protected (do not block)
[converting terms that a phrase negative could catch]

### Graduate to exact-match keywords
[converting search terms to promote]

### Cadence
Re-run this pass: [weekly / biweekly]

Worked example

• "plastic dog bowl" spent 14 USD, zero sales. Gate 1, drain, and gate 2, wrong material. Negate it. The token "plastic" is a whole irrelevant theme, so a phrase negative on "plastic" would be efficient if it is safe. Run the safety check: scan every converting term for the word "plastic". The winners here are "dog water bowl" and "stainless steel bowl", and neither contains "plastic", so the phrase negative cannot block a converter. Verdict: add "plastic" as a negative phrase. It also kills any future "plastic feeding bowl" or "plastic pet dish" before they spend. (Had a converting term contained "plastic", the rule flips: drop the phrase and negate "plastic dog bowl" as negative exact only.)
• "dog water bowl" converting at 4 USD per sale. Gate 3, winner. Keep and graduate to an exact-match keyword.
• "bowl" with 2 clicks, no data. Gate 4, noise. Wait.

Quality check

• The drain threshold is computed from the target cost per acquisition, not guessed.
• Every term is classified into one of the four buckets.
• No negative phrase is added without scanning for converting terms it would block.
• Converting search terms are graduated to exact-match keywords, not just kept.
• Low-click terms are left alone, not negated on noise.
• A recurring cadence is set.

Common mistakes

• Negating on noise. Blocking a term with 3 clicks and no sale. that is not data.
• Broad phrase negatives. Adding a negative phrase that quietly kills converting terms containing the same word.
• Negating winners. A high-ACoS term that still converts may be worth keeping at a lower bid, not negating.
• Find and forget. Running the pass once. new junk terms appear every week.
• Not graduating. Negating the junk but never promoting the converting discoveries.

Built by Jay GPT Pro