DataForSEO Brain

Play: Keyword Research Workflow

4 min read

Play: Keyword Research Workflow

Seed terms to a prioritized keyword list: Keywords Data for ground-truth volume, then DataForSEO Labs for expansion, difficulty, and intent. Sits under DataForSEO Brain → Flows.

Overview

This workflow expands a handful of seed keywords into a scored, intent-tagged target list. It blends two data lineages: the Keywords Data API (Google Ads-sourced volume/CPC/competition) for accuracy on a small set, and DataForSEO Labs (a pre-indexed in-house database of 8B+ keywords) for cheap, large-scale ideation, difficulty, and intent. Labs is the workhorse because it is far cheaper per query than live SERP.

Trigger

A topic, product, or page needs a keyword universe: a new content cluster, a PPC build, or a site launch. Inputs are 1-200 seed keywords plus a location_code and language_code.

Endpoints used (in order)

  • POST /v3/keywords_data/google_ads/search_volume/live (ground-truth volume/CPC for the seeds).
  • POST /v3/dataforseo_labs/google/keyword_ideas/live (category-relevant expansion).
  • POST /v3/dataforseo_labs/google/keyword_suggestions/live and /related_keywords/live (phrase and “searches related” expansion).
  • POST /v3/dataforseo_labs/google/bulk_keyword_difficulty/live (KD for up to 1,000 keywords/request).
  • POST /v3/dataforseo_labs/google/search_intent/live (intent for up to 1,000 keywords/request).
  • POST /v3/dataforseo_labs/google/keyword_overview/live (consolidated metrics for a final shortlist).

Pipeline

  1. Validate the seeds (UTF-8, lowercase) and resolve location_code + language_code.
  2. Pull anchor volume: send up to 1,000 seeds to Keywords Data search_volume (Google Ads lineage). Read search_volume, competition/competition_index (0-100), cpc, and monthly_searches[] for seasonality. One request covers 1 or 1,000 keywords at the same price.
  3. Expand with Labs keyword_ideas (up to 200 seeds, returns category-relevant terms with keyword_info volume/CPC/competition, keyword_properties.keyword_difficulty, and inline search_intent_info). Use closely_variants, ignore_synonyms, and filters to shape the set; page with offset_token beyond 10,000 results.
  4. Deepen with keyword_suggestions (long-tail phrase-match around a seed) and related_keywords (the “searches related to” graph) to capture demand the ideas endpoint misses.
  5. Score difficulty in bulk: batch up to 1,000 candidates into bulk_keyword_difficulty to get the 0-100 logarithmic KD per term in one billed task.
  6. Tag intent: send up to 1,000 keywords to search_intent (language-only) for keyword_intent.label (informational/navigational/commercial/transactional) plus probability and secondary_keyword_intents[].
  7. Prioritize: combine volume x intent x KD x CPC into a score, drop irrelevant or out-of-language terms (keyword_properties.is_another_language), and run keyword_overview on the shortlist for a clean consolidated record per keyword.
  8. Hand the scored list to play-content-strategy-brief or a PPC build.

Cost & cadence

  • Labs ideation/expansion endpoints bill about 0.0105).
  • Labs analytical endpoints bill about 0.0011 (cost-log: $0.0011).
  • Keywords Data Google Ads endpoints bill per request regardless of keyword count; the cost-log shows search_volume and keywords_for_site at $0.075 each.
  • Cadence: run on demand per topic. Volume/difficulty/intent change slowly, so a monthly refresh is plenty; Labs is Live-only so results return in roughly 1 second.

Output

A prioritized keyword table: keyword, search_volume, monthly trend, competition/CPC, keyword_difficulty, main_intent, and a composite priority score, ready for clustering, briefs, or campaign builds.

Pitfalls / limits

  • Keywords Data Google Ads is capped at 12 requests/min/account and returns combined volume for groups of similar keywords; certain emojis/UTF symbols are rejected.
  • Labs include_clickstream_data: true doubles the request cost; only enable when you need demographic/clickstream volume.
  • keyword_ideas takes max 200 seeds; keyword_suggestions/related_keywords take a single seed; do not over-batch.
  • Labs volume is a pre-computed estimate, not live Google Ads data; for ad-bidding accuracy confirm the final set against Keywords Data search_volume.
  • search_intent is language-only (no location) and Google-specific; there is no marketplace equivalent.

Decisions in play

  • dec-labs-vs-live-apis: route bulk expansion and difficulty/intent to pre-indexed Labs, reserve Keywords Data live volume for the anchor seeds and the final shortlist.
  • dec-which-api-for-which-job: Keywords Data for Google Ads ground-truth volume; Labs for ideation, KD, and intent at scale.
  • dec-cost-control-strategy: batch up to 1,000 keywords per Keywords Data request (flat price) and avoid include_clickstream_data unless demographics are needed.

Sources

Graph View

Backlinks