What the Keyword Planner API Actually Tells You (And What It Hides)
Keyword research gets much better when you understand what Google's API gives you, and what it quietly withholds.

The problem
Keyword research for a new SaaS product is a guessing game unless you can quantify demand. The Google Ads interface gives you estimates, but you can't automate it or integrate it into a strategy workflow. You need the API — specifically, KeywordPlanIdeaService — to pull structured search volume data programmatically and shape a paid search budget against real numbers.
The approach
The KeywordPlanIdeaService.generate_keyword_ideas() method takes seed keywords and a geographic/language targeting spec and returns monthly search volume, competition index, and low/high CPC bid estimates. The key non-obvious detail is that the volume figures are bucketed — Google returns a range (LOW, MEDIUM, HIGH) not a raw number, and the exact monthly figure is a different field (avg_monthly_searches) that requires you to specify the KeywordPlanIdeaServiceClient with the right customer_id from your MCC hierarchy, not a child account ID. Getting this wrong silently returns data for the wrong account's historical context.
The competition index is a float from 0.0 to 1.0 representing advertiser density in the auction. It is not a proxy for difficulty — a high-competition index just means a lot of advertisers are bidding, not that you can't win. CPC estimates are more useful: the spread between low (low_top_of_page_bid_micros) and high (high_top_of_page_bid_micros) tells you how volatile the auction is. A tight spread means stable pricing; a wide spread means aggressive bidding wars or thin auction depth.
For a campaign targeting "alternative to [competitor]" intent, I pulled three seed clusters across 30 target terms and found that the most commercially interesting terms were consistently mid-competition (0.4–0.6) with stable CPC spreads — the sweet spot for a new entrant with a controlled budget.
What I learned
The API is most useful not for the raw numbers but for the structural relationships between terms. Volume tells you what people search; competition tells you who's bidding; CPC spread tells you how sure the market is of the value. A term with high volume, low competition, and a tight CPC spread is a structurally underserved niche. Those are the terms worth building campaign architecture around — not the obvious head terms where incumbents have established quality scores and you're fighting uphill from day one.
