---
name: seo-keyword-strategy
description: Research keywords for a website or product and decide which pages to create, update, merge, or defer. Use when the user needs keyword research, search intent analysis, keyword clustering, or an SEO page plan backed by Toolmu DataForSEO data.
---

# Keyword and page strategy

## Before you start

Read [the shared research contract](../connect-toolmu/references/research-data.md) and [marketplace usage](../connect-toolmu/references/marketplace-usage.md). Use the existing Toolmu connection. The agent performs analysis; Toolmu supplies the evidence.

## Inputs

A website URL or product description, target country and language, audience, and conversion goal. Start from the website before asking for seed keywords.

Ask only for information that changes the task. Accept ordinary links and names; resolve technical identifiers from verified results. Confirm missing market or scope before dependent paid calls.

## Tool selection

Inspect the current contract with Toolmu `get_tool`, then call `execute_tool` only for the steps that need data:

- `onpage-content-parsing-dataforseo`
- `google-keyword-suggestions-dataforseo`
- `google-keyword-overview-dataforseo`
- `google-bulk-keyword-difficulty-dataforseo`
- `google-organic-serp-search-dataforseo`
- `google-ranked-keywords-dataforseo`
- `google-relevant-pages-dataforseo`

The catalog's current input and credit ceiling override examples in this document. Do not fabricate parameters, silently use a different supplier, or treat a search snippet as full source content.

## Workflow

1. Read the home page and up to two relevant product pages. List the actual product, audience, actions it supports, and existing landing pages. Treat unverified product claims as questions.

2. Form three to five seed topics from that inventory. Use keyword suggestions for the best two seeds first; normalize and deduplicate exact phrases while retaining the original wording.

3. Get volume and difficulty only for candidates relevant to the product. Start with at most 20 candidates. Read the live input contract before batching: a singular keyword field is not a batch endpoint. Never silently substitute a different market.

4. Inspect the top organic results for at most three leading queries. Record URLs, page types, titles, date, location, and language. Read relevant pages before judging their content; snippets alone do not establish a gap.

5. Cluster by shared intent and overlapping ranking URLs, then review meaning and the product fit. Do not group only by shared words or force an arbitrary overlap threshold. Assign one primary query to each page.

6. Map clusters onto the existing inventory. Choose create, update, merge, verify, or defer with a reason. An absent keyword in a limited ranking sample is not proof the site has no matching page.

7. Rank the first three actions by business relevance, reachable SERP format, evidence strength, and implementation effort. Present these dimensions separately rather than inventing a precise opportunity score.

## Deliver

A page-plan table: action, existing/proposed URL, audience task, primary and supporting queries, observed volume/KD with market and date, SERP evidence, proposed outline, conversion action, and reason for priority. Include a separate verify/defer list and the raw evidence ledger.

Include source links, actual market/date, returned sample size, missing coverage, run IDs, and confirmed credits. Let the user's agent choose the presentation; do not invoke an MCP App or a render tool.

## Boundaries

Do not promise ranking or traffic. A low KD does not establish weak product competition. Missing metrics stay unknown. Existing Toolmu market or batch restrictions must be disclosed before they affect the plan.

## Example request

> Plan SEO pages for [website URL] in [United States, English], focused on [product sign-ups].
