Google Trends & Autocomplete Keyword Research (12 Countries)

Type one keyword. For every country you choose, get:

• Autocomplete tails from the search boxes people actually use there — Google, Bing, YouTube, Amazon (US/JP), Naver (KR), Yahoo! JAPAN (JP), Baidu (CN)
• Buyer-intent breakdown of those tails: free, buy, bulk, howto, compare, tool, other — with up to five examples each
• Google Trends 12-month interest for the keyword in that country (optional)

Use it before you build or stock something: a tail full of "free" means people want it free; tails with "price", "best", "bulk", "批量", "一括", "일괄" mean someone is ready to pay.

Use cases

• Before you build a product or tool — see whether people in a country add "free", "price" or "bulk" after your keyword.
• Picking a market — run the same idea in each country's own language and compare the intent mix (not the Trends numbers).
• Content and SEO briefs — the howto and compare tails are ready-made article and video topics.
• Marketplace sellers — Amazon (US/JP) and Naver/Yahoo! JAPAN tails show how shoppers phrase the product.

Example output (shortened)

{
  "keyword": "invoice template",
  "country": "US",
  "suggestionCount": 23,
  "intentShare": {"free": 0.348, "buy": 0.0, "bulk": 0.0, "howto": 0.0, "compare": 0.0, "tool": 0.043, "other": 0.609},
  "topBuyTails": [],
  "trends": {"first4w": 69.0, "last4w": 46.5, "mean12m": 73.7, "scale": "relative 0-100 within this country; not comparable across countries"}
}

Real run on 2026-10-07. Reading: a third of US tails ask for "free" and none show buying words — a free tool or lead magnet fits better than a paid template here.

FAQ

Is this search volume? No. It tells you what people type and what they want, not how many. Pair it with a volume tool if you need counts.

How much does a run cost? One keyword in three countries = 3 results = $0.015 on the free plan. Paid Apify plans get Store discounts: $4.50 (Starter), $4.00 (Scale) and $3.50 (Business) per 1,000 results. Pairs where every source failed are free.

Is there a ready-made example? Yes — Keyword demand in English-speaking markets runs two keywords across the US, UK, Canada, Australia and India.

Can I schedule it? Yes — use Apify Schedules and get a fresh snapshot every week.

Input

{
  "keywords": ["invoice template", "請求書 テンプレート"],
  "countries": ["US", "JP"],
  "includeTrends": true
}

Write each keyword in the language people search in that country. Supported countries: US, GB, CA, AU, IN, KR, JP, DE, FR, ES, BR, CN.

Output (one item per keyword × country)

Field	Meaning
keyword, country	What was checked
suggestionCount	Unique autocomplete tails across sources (the seed itself excluded)
intentCounts, intentShare	Tails per intent, and share of all tails. A tail can match more than one intent
intentExamples, topBuyTails	Example tails per intent
suggestions	Raw tails per source
sourceErrors	Sources that failed this time. Failures are shown, never hidden
trends	first4w, last4w, mean12m on Google's 0-100 scale within that country — do not compare countries with it. Includes the Google Trends attribution

Pricing

Pay per result: one keyword-signal event per keyword × country. Apify also adds its standard actor start fee of $0.00005 per run (once per GB of memory; default runs use 256 MB), which covers the first seconds of compute. You are not charged when every autocomplete source fails for that pair. The run stops when your spending limit is reached.

Limits — read before you rely on it

• Autocomplete shows what people type, not how many. There are no search volumes here.
• Google Trends values are relative and may be rate-limited; when that happens the field says so.
• Intent words are a fixed multilingual list. Tails it does not recognise fall into other.
• Sources are public search-box suggestion endpoints and can change without notice.

Data source for trend values: Google Trends (https://www.google.com/trends).

Not affiliated with, endorsed or sponsored by Google, Microsoft, Amazon, Naver, LY Corporation (Yahoo! JAPAN) or Baidu. Their names are trademarks of their owners, used here only to describe the public data sources.