Workflow: Turn competitor wins into playbooks
After WF1 identifies the fastest-growing competitors, this step zooms in on the one you select for a full keyword cluster breakdown — which topics they've built an organic footprint in, how many pages they've built, and where their traffic concentrates.
Add your data
Adding details ensures a relevant result
Save time by automatically applying details and getting ready-to-use prompts.
Prompt
Using Semrush organic data for {country}:
Analyze {competitor-domain} if provided.
If a previous competitor analysis exists in this conversation, also consider the top growing competitors identified there. Prioritize {competitor-domain} first if provided, then include up to 3 additional growing competitors from the previous analysis if they are relevant.
If {competitor-domain} is not provided and previous growing competitors are available, select the most relevant growing competitor from the previous analysis.
If neither {competitor-domain} nor previous competitor analysis is available, ask the user to provide a competitor domain before proceeding.
1. Pull the organic keywords {competitor-domain} ranks for (keyword, position, monthly volume, URL). Display_limit: 100. Sort by volume descending.
2. Group the returned keywords into 5 topic clusters by semantic similarity. Infer cluster labels from keyword patterns (e.g., "backlink analysis", "site audit", "keyword research"). Label all cluster names as (inferred).
3. For each cluster: count unique ranking URLs and identify the top 3 keywords by volume.
Return ONE table:
Columns:
* source_competitor
* cluster_name
* keyword_count
* unique_pages_count
* top_3_keywords (with volume)
* growth_signal (high/med/low)
Growth signal rule:
* high = top 2 clusters by keyword count
* med = clusters 3–4
* low = cluster 5
Limit: 5 clusters.
Complete the keyword pull before returning the table.
Label all cluster groupings as (inferred).
Example Output (illustrative)
- A 5-cluster breakdown for {competitor-domain} — keyword count, unique ranking pages, top 3 keywords with volume, and growth signal (high / med / low); cluster names labeled as inferred