# How to find where you pay for clicks you already rank for

> Tier: deep. Full contract: quantified decision rules and evidence requirements.

The queries where the site holds strong organic positions and buys ads anyway, sized in spend, so the incrementality conversation has numbers.

**Intent.** Quantify the overlap between paid and organic search, and separate defensible spend from probable substitution.

## When to use

- Brand campaigns consuming meaningful budget.
- CFO or client asks the classic question: why do we pay for our own name?
- After organic gains on commercial terms.

## When not to use

- Sites with weak organic presence: no overlap to find.
- As a substitute for an incrementality test: this sizes the question, only a holdout answers it.

## Sources

Required: Google Ads, Search Console

## Preconditions

- Both sources connected and ticked for the same site and market.
- 90 days of data on both for stable positions.

## Inputs

- `window`: Analysis window. Default: last 90 days.

## Procedure

1. Pull the combined paid and organic view from Google Ads. — tools: `google_ads_get_paid_organic_search_terms`
2. Pull organic positions and CTR by query from Search Console. — tools: `gsc_query_search_analytics`
3. Classify overlap queries: brand, category with position 1-3, category below, competitor terms. — tools: `gsc_cluster_queries`
4. Size paid spend per class over the window. — tools: `google_ads_get_search_terms`

## Decision rules

- **When** Non-brand queries with organic position 1-2 carry paid spend → **conclude** Highest substitution risk; size it and propose a holdout test, not an immediate cut.
- **When** Brand queries show competitor ads in the auction → **conclude** Defensive spend is justified; report the presence as the reason.
- **When** Organic position for an overlapped query is 4-10 → **conclude** Paid is doing real work there; keep, and note the SEO opportunity instead.

## Evidence requirements

- Every class carries query count, paid spend, average organic position, both-source clicks.
- No incrementality percentage is asserted anywhere: the skill sizes exposure, the test measures it.

## Human approval boundary

- Pausing brand or overlap spend is a strategy decision with revenue risk; the skill proposes the test design, humans decide.
- Any pause executed later goes through Safe Writes.

## Failure modes

- Search Console and Google Ads bucket queries differently; join on normalised queries and state the match rate.
- Seasonal organic volatility reshuffles positions; a 7-day read produces false substitution findings.
- Cutting brand spend shows organic clicks rising and total clicks falling in many tested cases; never promise the saving equals the spend.

## Output

- **Overlap map**: Classes, spend, positions, click totals.
- **Test design**: The holdout proposal: scope, duration, success metric.
- **SEO handoffs**: Queries where the cheaper fix is ranking better, not bidding.

## Related

- Skill: [keyword-expansion](https://www.getmcpads.com/skills/keyword-expansion.md)
- Skill: [organic-decay-triage](https://www.getmcpads.com/skills/organic-decay-triage.md)
- Guide: [search-console-analysis](https://www.getmcpads.com/guides/search-console-analysis.md)