Google Autocomplete Keywords - Intent & Changes

Turn seed phrases into a repeatable keyword-research watchlist. Get the exact suggestion, its position in each source response, the query that produced it, and an optional comparison with your last saved observation.

Start with five suggestions

{"seeds":["accounting software"],"maxResults":5}

Choose Start, then open Keyword opportunities or the readable report. Export CSV/JSON, or give the same input to an AI agent through Apify's API or MCP server. No source API key or paid proxy is required.

Built for a business workflow

• SEO teams: collect language-specific topics and question phrases for editorial review.
• Product marketers: watch literal pricing, comparison and alternative queries around a category.
• Agencies: save a task per client, schedule repeats and compare only the evidence that was actually observed.

Use Pricing and alternatives to add explicit pricing/alternatives/vs query variations. Use Question prefixes for how/what/best variations. The Actor shows every generated query; it does not hide extra source requests.

Field	Meaning
seed, suggestion	Original seed and source-returned phrase
queryEvidence	Each query, suggestion rank and exact source URL
intent	Literal rule-based alternative/comparison/pricing/question/general label
language, country	Requested language and country hint
recordId, observedAt	Stable identity and observation time
comparisonStatus, changedFields, previous	Saved observation comparison when enabled

Intent labels are organizational rules, not predictions of a person's intent. Suggestion ranks are not Google search-result rankings. Country is a hint; request location and Google's own systems can affect results. This Actor provides no search-volume, CPC, traffic or SEO-difficulty estimates.

Saved monitoring

Enable comparePrevious, choose a snapshotName, and reuse the same seeds, language, country, expansion and query cap. The first run establishes a baseline. onlyChanges returns only new or changed observed suggestions; an unchanged run can correctly save zero rows. History lives in your account. Only confirmed saved rows advance it. One writer per snapshot; do not overlap runs. Missing suggestions do not prove deletion or a trend reversal.

Coverage and controls

Up to ten seeds,40 source queries and500 saved suggestions per run. Queries run sequentially at a paced rate and share a30–180second deadline. Results are deduplicated per seed and phrase; evidence retains the contributing query ranks. Maximum results can cut output short. Check SUMMARY for requested/reached queries, caps and failures before treating an export as complete. Public autocomplete can change, return no suggestions or rate-limit access. A source failure produces diagnostics, never a paid fake keyword. Robots policies are checked. No login, proxy rotation, challenge solving or access-control bypass is used.

Pricing

Base price: $0.001 per saved suggestion, plus $0.00005 per start at default memory. Five suggestions cost $0.00505 at the base rate. Plan discounts and the Store pricing panel govern actual charges. Set a maximum run charge to cap spending. Failed source requests, reports and skipped unchanged suggestions are not billed result rows.

Three useful recipes

{"seeds":["accounting software"],"maxResults":5}

{"seeds":["crm software"],"expansion":"commercial","maxQueries":4,"maxResults":10}

{"seeds":["marketing software"],"language":"de","country":"DE","maxResults":5}

Create a task with your inputs and attach an Apify Schedule. Your own integrations can send the dataset to Sheets,n8n,Make or a data pipeline; no delivery is activated automatically.

How Google autocomplete works. This independent Mayd IT tool is not affiliated with Google. For help, open an Actor issue with a run ID and expected behavior, or contact bmay@mayd-it.com; never include credentials.