Free tool · Understand my results
Which paid-search queries should I review?
Review a local paid-search query export with your own labels and compatible-context totals.
Which paid-search queries should I review?
Use your own figures for deterministic planning or analysis. Inputs remain in this browser. Models do not verify business outcomes; contribution is not net profit.
How to use this tool
Bring: A CSV search-term export, explicit column mapping, separator and decimal convention. Supply currency, period and conversion definitions where absent.
- Use a local CSV export without personal or confidential data. Choose its separator and decimal convention explicitly.
- Map search term and metric columns, then confirm currency, period and reported conversion definitions.
- Review flags in context and assign your own labels. Sort and filter without assuming a flagged query is waste.
- Download the full owner-labeled CSV if needed. Make any account or negative-keyword decisions separately after match-type and conflict review.
Hypothetical example — fictional inputs, not a forecast
A fictional row for “boiler service” reports $120 spend, 30 clicks and 2.5 platform-attributed conversions under a stated USD period and definition.
The row retains 2.5 reported conversions rather than rounding it into unique customers. Labels remain owner decisions; no negative keyword is applied.
Use these figures only to understand the method. Replace them with your own consistent inputs.
Read the result in context
Compatible context matters more than a single number. Review overlap, conversion delay, meaning and business relevance before interpreting totals.
Assumptions and limitations
- Local CSV limit: 1 MiB, 5,000 data rows, 30 columns and 10,000 characters per cell.
- The display shows the first 200 matching rows or groups; the full retained rows remain exportable.
- Duplicate candidates are retained and flagged, not silently deduplicated; unknown contexts are not silently merged.
Your information
CSV files are read in your browser and are not uploaded by this tool. Labels remain in the session unless you explicitly download, copy or print. Remove personal data before loading an export. These statements describe the tool code. Other website scripts have their own privacy practices; review the website privacy policy before entering sensitive business information.
Questions before you start
Does this apply negative keywords?
No. It does not connect to an account, recommend automatic negatives or make account changes.
Why retain fractional conversions?
Some platform attribution models report fractional credit. The tool preserves the reported number and its definition rather than calling it a verified customer count.