DataForSEO Brain

Google Ads Keyword Planner (upstream source)

4 min read

Google Ads Keyword Planner (upstream source)

The Google Ads volume source that feeds DataForSEO keyword data - and the “bucketing” problem DataForSEO’s clickstream metric is designed to fix. Sits under DataForSEO Brain → Entities.

Overview

Google Ads Keyword Planner (GKP) is the upstream search-volume source behind cap-keywords-data-api. DataForSEO’s keyword index is “collected using Google Ads and Bing Ads services,” so GKP is a primary lineage point. All major SEO tools - DataForSEO, ent-ahrefs, ent-semrush, ent-moz - model search volume from Google Ads “buckets,” then refine differently. Understanding GKP’s behavior explains both where DataForSEO’s volume comes from and why a refinement layer exists.

The bucketing problem

  • Keyword Planner returns ~80 logarithmically distributed repeated values: many distinct queries get collapsed into the same coarse bucket (e.g. a shared “1K-10K” band).
  • This makes raw GKP volume poorly differentiated at the individual-keyword level.
  • Tool-to-tool volume discrepancies commonly run 30%-200%, driven by how each provider groups query variants and which panel it samples.

How DataForSEO refines it

  • DataForSEO exposes raw Google Ads volume directly, plus a distinct “DataForSEO Search Volume” metric: it takes Google Ads SV, groups all queries in a bucket, then re-distributes per-keyword using either Bing Ads SV or clickstream event counts.
  • The use_clickstream parameter selects the refined model; clickstream-refined endpoints are charged more (see ent-clickstream-data and cap-trends-and-clickstream).
  • Practitioner caveat: by default the plain volume is “generally an average value of what you see in Google Keyword Planner,” i.e. less differentiated unless you opt into the clickstream-refined endpoint.

When to use / how it fits

Gotchas / limits

  • Because clickstream panels are samples, not censuses, no provider (GKP-derived or otherwise) is “correct” in absolute terms.
  • Bing-normalized and clickstream-normalized fields (keyword_info_normalized_with_bing, keyword_info_normalized_with_clickstream) are separate from the raw Google Ads search_volume - pick deliberately.
  • Default endpoints can mask real per-keyword differences hidden inside a GKP bucket.

Lineage in DataForSEO fields

  • Raw Google Ads volume surfaces as keyword_info.search_volume with competition, competition_level, cpc, low_top_of_page_bid, and high_top_of_page_bid on Labs keyword items.
  • monthly_searches carries the trailing twelve months; search_volume_trend reports monthly/quarterly/yearly percent change.
  • Bing-normalized and clickstream-normalized variants live in separate objects, so you can compare GKP-derived volume against refined estimates side by side.

Practitioner guidance

  • Use raw GKP volume for directional sizing and ad-cost signals (CPC, bids).
  • Switch to a normalized model when ranking decisions hinge on differences between keywords inside one GKP bucket.
  • Validate against your own Search Console impressions where possible, since no panel is a census.

Why a refinement layer exists

  • GKP is built for advertisers sizing ad reach, not SEOs ranking individual terms, so its native granularity is coarse by design.
  • Re-distributing a bucket with Bing Ads volume or clickstream events recovers per-keyword differences GKP collapses.
  • This is the same problem every SEO tool faces; DataForSEO’s answer is to expose both the raw and the refined number and let you choose.

Cost signal (this brain’s runs)

  • keywords_data/google_ads/search_volume/live cost 0.075 for ~1,921 returned results.
  • Google Ads keyword calls are roughly an order of magnitude pricier than equivalent Labs lookups (~$0.0105), so route deliberately.
  • Bing Ads volume (keywords_data/bing/search_volume) is a parallel upstream source and was observed at $0.075 per call, the same tier as Google Ads volume.
  • Because both Google Ads and Bing Ads feeds carry this cost tier, push high-volume keyword discovery to the pre-indexed Labs engine and reserve the Ads endpoints for authoritative spot checks.

Sources

Graph View

Backlinks