People Also Ask Scraper

What is People Also Ask Scraper?

People Also Ask Scraper extracts the Google "People also ask" questions for any list of keywords. One row is one question, with the keyword it came from, its nesting level (shown on the page or revealed by expanding another question), the parent question that revealed it, its position in the box, the country, language and device of the search, and a link to the Google results page. Optionally it also returns Google's Related searches and People also search for suggestions.

The data comes from live Google results pages in any country and language Google serves, on desktop or mobile, through a licensed SERP data provider. No Google account, no API key, no proxy setup and no official Google API needed. Questions cost $0.0009 each ($0.90 per 1,000), there is no fee per run, and keywords with no "People also ask" box are free.

Use it as a PAA scraper for content ideas, FAQ schema research, blog outlines, topic clusters and AI content briefs — a Google questions API you can call from Python, JavaScript, n8n, Make or an AI agent.

What data can People Also Ask Scraper extract?

Field	Example	What it means
type	question	Row type: question, related-search (only when switched on) or error (free)
keyword	how to start a podcast	The keyword you searched
question	Do I need a license for a podcast?	The question exactly as Google shows it
level	2	1 = on the results page, 2 = appeared after Google expanded a level-1 question, 3–5 = deeper expansions
parentQuestion	How to start a podcast with no audience?	The question whose expansion revealed this one (empty for level 1)
position	6	Order of the question in the box, starting at 1
answerType	ai_overview_not_captured	How Google answers the question (see "Answers" below)
serpPosition	3	Where the "People also ask" box sits among all elements of the results page
location, locationCode	United States, 2840	Country of the search
language, languageCode	English, en	Google interface language
device	desktop	desktop or mobile
searchUrl	https://www.google.com/search?q=…	The Google results page, to check any row by hand
scrapedAt	2026-10-01T14:23:33Z	When the row was produced

Related-search rows (only when you switch them on) carry keyword, query (the suggested search), source (related_searches or people_also_search), position, country, language, device and searchUrl.

🧭 Questions as a tree, not a flat list

level and parentQuestion let you rebuild the question tree Google builds when someone clicks through the box. In our test run for how to start a podcast (United States, English, desktop, depth 2) Google showed four questions on the page; expanding "How to start a podcast with no audience?" added "Why do most podcasts fail?" and "Do I need a license for a podcast?", and expanding the licence question added three more, including "Do podcasters have to pay their guests?". That is how you see which sub-topics Google connects to your keyword — the structure of a good FAQ section or a pillar page.

Answers: what Google shows in 2026

Google used to answer each "People also ask" question with a quoted paragraph from a web page. In 2026 it answers most of them with an AI-generated answer that loads after the page. Our data provider does not capture those AI answers, so on such questions answerType is ai_overview_not_captured and there is no answer text. In our October 2026 test runs (United States and Germany, desktop and mobile) this was the case for every one of the 56 questions we checked.

When Google does show a classic quoted answer, the row also carries answer, sourceUrl, sourceDomain and sourceTitle, and answerType is featured_snippet (or ai_overview when an AI answer was captured). Those fields simply do not appear on rows without an answer. The price per question is the same either way: you pay for the question.

How much does it cost to scrape People Also Ask questions?

People Also Ask Scraper uses pay-per-event pricing. You pay only for rows with data.

Event	Price (free plan)	What you get
question	$0.0009 ($0.90 per 1,000)	One People Also Ask question row
related-search	$0.0005 ($0.50 per 1,000)	One Related searches / People also search for row — only when you switch it on
Run start	$0	No fee per run
Keyword without a People Also Ask box	$0	Free row with hasData: false
Error rows	$0	Bad country, bad language, empty keyword — never charged

Paid Apify plans pay less per question: Bronze $0.00085, Silver $0.0008, Gold and above $0.00075.

Set Max cost per run in the run options and the Actor stops exactly there: it never hands out rows past your limit, and it does not keep searching once the limit is reached.

Real numbers from our test runs:

• The default input (2 keywords, depth 2) returned 18 questions for $0.0162.
• One German keyword with Related searches on returned 9 questions and 8 related searches: $0.0121.
• One keyword on mobile at depth 0 with People also search for on: 4 questions + 10 suggestions = $0.0086.

A typical keyword at the default depth gives about 9 questions, so $1 buys roughly 1,100 questions, or about 120 keywords fully expanded. Apify's free plan includes $5 of monthly usage, enough for about 5,500 questions.

💰 How expansion depth changes the cost

Each keyword is one Google search. Question expansion depth tells Google how many times to open a question and collect the new ones it adds. More depth = more questions per keyword = more rows billed:

Depth	Questions per keyword (measured)	Cost per keyword
0	4	about $0.0036
1	6	about $0.0054
2 (default)	9	about $0.0081

Depths 3 and 4 continue the same way (Google adds two to three questions per expansion). The number of questions is Google's, not ours: some keywords show fewer, and some show none.

How to scrape Google People Also Ask with People Also Ask Scraper

1. Open People Also Ask Scraper in Apify Console and click Try for free.
2. In 🔍 Keywords, type one search phrase per line — the way people search, e.g. bali visa, how to start a podcast.
3. Set 📍 Country and 🗣️ Language (names or codes: United States, GB, Germany, de, Indonesian, id).
4. Choose ❓ Question expansion depth (default 2) and the 💻 Device.
5. Switch on 🔗 Add Related searches or 🔗 Add People also search for if you want those suggestions too.
6. Click Start. Most runs finish in under a minute; repeating the same keywords within 7 days is answered from cache in a few seconds.
7. Download the results as CSV, Excel, JSON or HTML, or read them over the API.

⬇️ Input

The default input runs two keywords in the United States, in English, on desktop, expanded to depth 2:

{
  "keywords": ["bali visa", "how to start a podcast"],
  "location": "United States",
  "language": "English",
  "clickDepth": 2,
  "device": "desktop",
  "includeRelatedSearches": false,
  "includePeopleAlsoSearch": false,
  "maxKeywords": 100
}

Field	Default	Notes
keywords	2 demo keywords	Up to 1,000 per run, one Google search each
location	United States	Country name or two-letter code; every Google country is supported
language	English	Language name or code; empty = the country's main language (Indonesia → Indonesian, Vietnam → Vietnamese, Germany → German)
clickDepth	2	0–4; more depth = more questions ($)
device	desktop	desktop or mobile
includeRelatedSearches	false	Adds Related searches rows ($)
includePeopleAlsoSearch	false	Adds People also search for rows when Google shows them, mostly on mobile ($)
maxKeywords	100	Safety cap on the number of keywords searched

Questions only (the default)

Leave both 🔗 switches off. The dataset then has one row type — question — and the Questions tab shows a clean table.

Questions plus Related searches

Switch on 🔗 Add Related searches. Each keyword then also returns its Google Related searches (8 per keyword in our tests) as related-search rows. They appear in the Related searches tab; type is the first column in every tab so the two row types never mix up.

⬆️ Output

One question row from the default run:

{
  "type": "question",
  "keyword": "bali visa",
  "question": "Is it safe for Americans to travel to Bali right now?",
  "hasData": true,
  "level": 2,
  "parentQuestion": "Should I get a Bali visa online or on arrival?",
  "position": 6,
  "answerType": "ai_overview_not_captured",
  "serpPosition": 3,
  "location": "United States",
  "locationCode": 2840,
  "language": "English",
  "languageCode": "en",
  "device": "desktop",
  "searchUrl": "https://www.google.com/search?q=bali%20visa&hl=en&gl=US&ie=UTF-8",
  "scrapedAt": "2026-10-01T14:23:33.088Z"
}

One related-search row:

{
  "type": "related-search",
  "keyword": "podcast starten",
  "query": "Podcast starten kostenlos",
  "source": "related_searches",
  "position": 2,
  "location": "Germany",
  "language": "German",
  "device": "desktop"
}

A keyword with no "People also ask" box returns one free row with question: null, hasData: false and a note explaining it, so your list stays complete and you can see which keywords Google does not attach questions to.

The run summary (keywords searched, questions found, keywords without a box, rows from cache) is saved in the key-value store as SUMMARY.

Use cases for Google People Also Ask data

Content ideas and blog outlines

Every question is a sub-heading someone actually searches. Run your topic list, sort by keyword and level, and you have the H2/H3 skeleton of an article that answers what readers ask next.

FAQ schema research

Pick the level-1 and level-2 questions for a page's keyword and answer them in a FAQ block. The questions are Google's own wording, which is what you want in FAQ markup and in AI search answers.

Topic clusters and pillar pages

parentQuestion shows which questions Google treats as follow-ups of which. Questions that branch from the same parent belong on the same page; separate branches are candidates for separate cluster articles.

Keyword research pipelines

Chain it with our other SEO actors: get long-tail keywords from Keyword Suggestions Scraper, pull the questions for each with People Also Ask Scraper, then check which questions have search demand in Keyword Search Volume Checker.

AI content briefs and agents

A writing agent can call the Actor through MCP with a keyword, get flat question rows back and turn them into an outline — no HTML parsing, no Google blocking.

Local and international SEO

The same keyword in different countries and languages produces different questions. Run bali visa for the United States, the United Kingdom and Australia and compare what each market wants to know.

Integrations

Apify API

curl -X POST "https://api.apify.com/v2/acts/lergassy~people-also-ask-scraper/run-sync-get-dataset-items?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"keywords": ["how to start a podcast"], "location": "US", "language": "en"}'

Python

from apify_client import ApifyClient

client = ApifyClient("YOUR_TOKEN")
run = client.actor("lergassy/people-also-ask-scraper").call(run_input={
    "keywords": ["how to start a podcast", "podcast microphone"],
    "location": "United States",
    "language": "English",
    "clickDepth": 2,
})
for row in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(row["level"], row["question"])

JavaScript

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: 'YOUR_TOKEN' });
const run = await client.actor('lergassy/people-also-ask-scraper').call({ keywords: ['bali visa'], location: 'GB' });
const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items.map((r) => r.question));

No-code tools

Use the Apify integrations for Google Sheets, n8n, Make, Zapier, Slack and webhooks to send questions to a spreadsheet or a content calendar after every run, or schedule a weekly run to track how the questions for your keywords change.

MCP server for AI agents

The Actor works with the Apify MCP server (https://mcp.apify.com), so Claude, ChatGPT, Cursor and other MCP clients can call it as a tool.

🤖 For AI Agents & LLM Apps

Compact reference for agents calling lergassy/people-also-ask-scraper through the Apify MCP server or API.

Purpose: returns the Google "People also ask" questions for keywords; one row per question with its nesting level and parent question.

Minimal input:

{ "keywords": ["how to start a podcast"], "location": "US", "language": "en" }

Output fields (question rows): type, keyword, question, hasData, level, parentQuestion, position, answerType, serpPosition, location, locationCode, language, languageCode, device, searchUrl, scrapedAt; plus answer, sourceUrl, sourceDomain, sourceTitle only when Google showed a quoted answer. Related-search rows: type, keyword, query, source, position, location, language, device, searchUrl.

Behaviors an agent should know:

• One keyword = one Google search; clickDepth (0–4, default 2) controls how many extra questions are revealed.
• Answers are usually not available (answerType: ai_overview_not_captured); do not expect answer text.
• hasData: false rows mean Google showed no question box for that keyword; they are free.
• Related-search rows appear only when includeRelatedSearches or includePeopleAlsoSearch is true; filter by type.
• Billing: question $0.0009, related-search $0.0005; errors and empty keywords are free; no run fee.
• Rows are flat; no nested objects.

❓ FAQ

Is it legal to scrape Google People Also Ask?

The Actor reads public Google search results, the same page anyone sees without logging in, through a licensed SERP data provider. It collects no personal data. Using public search data for research and content planning is common practice, but rules differ by country and by use — consult a lawyer if you are unsure about your case.

Does it return the answers to the questions?

Only when Google shows a classic quoted answer. In 2026 Google answers most "People also ask" questions with an AI answer that loads after the page, and that text is not captured — in our tests, none of 56 questions had a capturable answer. Treat this Actor as a question scraper; you are billed per question, not per answer.

How many questions do I get per keyword?

Usually 4 on the page; about 6 at depth 1 and about 9 at depth 2 in our tests. Some keywords show fewer, and some show no box at all — those return a free hasData: false row.

Why are the questions different from what I see in my browser?

Google personalises results by location, language, device and history. The Actor searches as a clean, logged-out user in the country and language you choose. Open searchUrl to see the same page.

Can I use People Also Ask Scraper with the Apify API?

Yes. Every run can be started and read over the Apify REST API, or with the Python and JavaScript clients shown above.

Can I use it through an MCP server?

Yes. Connect the Apify MCP server to Claude, ChatGPT, Cursor or any MCP client and call lergassy/people-also-ask-scraper as a tool.

Can I integrate it with other apps?

Yes — Google Sheets, n8n, Make, Zapier, Slack, webhooks and any tool that reads JSON or CSV.

How much does it cost to scrape 1,000 keywords?

At the default depth about 9,000 questions, roughly $8.10. At depth 0 about 4,000 questions, roughly $3.60. Keywords without a question box cost nothing.

Is one row one question?

Yes. Every question is its own row with its keyword, level and parent, so you can sort, filter and count without unpacking nested data.

What happens if many keywords return nothing?

The Actor stops early and writes one free error row if the keywords keep returning no question box — for example, a list of random strings. Rows already delivered stay in the dataset.

Your feedback

Found a bug, missing a field or need another country or device setting? Open an issue in the Issues tab — we read every one. If the Actor saves you time, a review on the Store page helps other people find it.

You might also like

Actor	What it does
Keyword Suggestions Scraper	Google, YouTube, Amazon, Bing, eBay and TikTok autocomplete suggestions — the long-tail keywords to feed into this Actor
Keyword Search Volume Checker	Monthly search volume, CPC, difficulty and intent for your keywords and questions
Google Trends Scraper	Interest over time, by region, related queries and topics
Competitor Keywords Finder	Every keyword a competitor domain ranks for on Google