KOF (Keyword Opportunity Finder) — Apify README

What it does (high level)

• Expands seeds into candidate keywords using SerpApi (titles, People Also Ask, related searches).
• Fetches metrics (MSV/CPC from DataForSEO Google Ads, KD from DataForSEO Labs) and filters out keywords without MSV.
• Clusters related queries, picks a primary keyword per cluster, and takes a quick SERP snapshot for context.
• Scores opportunity (volume up, difficulty down, small bonuses/penalties from SERP features/brands).
• Generates a concise brief for the primary keyword (title/H2/FAQs/meta) via OpenAI.

Who is this for?

• [Content and SEO teams] looking for fast, high-signal topic discovery and prioritization.
• [Product-led growth and marketing] evaluating where to invest content for capture/convert.
• [Studios, agencies, and builders] who need a reusable, automatable keyword opportunity engine.

What business value does it provide?

• [Prioritization] Surfaces groups of related keywords with an opportunity score to guide where to publish next.
• [Speed] Generates concise, production-ready briefs per primary keyword to accelerate drafting.
• [Context] Includes SERP indicators that hint at competitiveness and angles to differentiate.
• [Repeatability] Deterministic runs and pagination make it easy to operationalize.

Can it work with other AI agents and automations?

• [Yes]. Results are saved to OUTPUT.json (and can be mirrored to a dataset) so you can plug this into:
  • n8n flows (e.g., auto-create tasks in Notion/Jira, or kick off drafting pipelines).
  • LangGraph / LangChain agents (consume clusters and briefs as inputs to writing/research agents).
  • Zapier / Make for low-code automations (publish briefs to CMS, notify teams, etc.).

How to use (Apify)

• Run the Actor from Apify Console and configure your run (API keys, high-level options). Click Start.
• You’ll find results in the run’s storage: Key‑Value Store under OUTPUT.
• A short, human‑readable summary is printed in logs at the end of each run.

Output

• [Where] storage/key_value_stores/default/OUTPUT.json
• [What] Clusters with opportunity score, SERP context, and optional content briefs; metadata and a concise summary string.

Example OUTPUT.json (abridged)

{
  "fingerprint": "<run-hash>",
  "cursor": null,
  "stats": {
    "seeds": 1,
    "keywordsFetched": 6,
    "clusters": 2,
    "serpCalls": 3,
    "metricsCalls": 2,
    "llmCalls": 2
  },
  "clusters": [
    {
      "id": "c9abc2e755743f5babfb1e31e862d5b260f83f45",
      "primary_kw": "n8n ai",
      "members": ["n8n ai", "n8n automation"],
      "metrics": {
        "avg_volume": 1320,
        "avg_cpc": 6.04,
        "avg_kd": 27,
        "total_volume": 2640
      },
      "serpSnapshot": {
        "topDomains": ["n8n.io", "n8nchat.com"],
        "features": ["paa"]
      },
      "opportunity": 61,
      "brief": {
        "title": "Unlocking N8n AI for Automation Success",
        "h2": [
          "What is N8n AI?",
          "Benefits of Using N8n AI",
          "How to Get Started"
        ],
        "faqs": [
          "Q: What is N8n AI? A: An open-source automation tool integrating apps/services.",
          "Q: Is it free? A: Yes, with paid hosting/support options."
        ],
        "metaTitle": "N8n AI: Revolutionize Your Automation",
        "metaDescription": "Discover how N8n AI transforms automation and productivity."
      },
      "resultHash": "<cluster-hash>"
    }
  ],
  "newClusters": [/* same page subset */],
  "summary": "✅ 2 clusters found (avg opportunity 45)\n🔥 Best cluster: \"n8n ai\" (score 61)\n📄 2 briefs generated\n---"
}

Costs (high‑level)

• [SerpApi] Requests scale with seeds and the number of clusters processed per page.
• [DataForSEO] We batch keyword metrics; each request can include up to 1,000 keywords (configurable). Observe provider rate limits when increasing batch size.
• [OpenAI] Briefs use gpt-4o-mini (cheap). About 300 tokens per request.

You’ll see request counts and brief counts in the stats field of OUTPUT.json, which helps back-calculate spend.