People Also Ask by country: the questions barely overlap, and the US is the odd one out

People Also Ask is one of the best free sources of questions people really have. If you sell in more than one market, the obvious shortcut is to collect it once, usually for the US, and reuse it. We tested whether that works.

We ran 10 searches in five countries and compared the questions. This guide has the numbers, a function that does the comparison for your own seeds, and a result we did not expect: the US is the odd one out. We used the Google Search API in October 2026. The box changes over time, so yours will differ.

How we ran it

We used 10 informational searches: how to lower blood pressure, what is inflation, how to start a podcast, best credit card, how to learn python, how to boil eggs, why is the sky blue, how to change a tire, what causes headaches and how does a vpn work. Each ran in us, gb, au, ca and in, in English, on desktop. That is 50 requests. One, how to change a tire in India, failed after retries with Google temporarily blocked the request. Please retry., so India has 9 searches.

It is a small set of seeds we picked, checked once, so the figures describe it and nothing wider.

The box is always there, and always 4 long

Every search that completed had a People Also Ask box, in every country: 10 of 10 in the US, UK, Australia and Canada, and 9 of 9 in India. Every box had exactly 4 questions, so the 49 searches produced 196 questions, of which 116 were different.

The US shares almost nothing with the others

For each pair of countries, we counted how many of the 4 questions two boxes had in common, on average across the seeds:

Look at the pattern. The UK, Australia, Canada and India share 2.25 of 4 questions with each other on average. With the US, they share 0.28. Counting questions rather than pairs:

• 37 of the 40 US questions appeared in no other country.
• 36 questions appeared in two or more non-US countries but not in the US.
• Only 2 questions appeared in all five countries, and 77 of the 116 different questions appeared in exactly one country.

A concrete case: what is inflation returned What is inflation in brief? in the UK, Australia and Canada, What is the current real inflation rate in the United States? in all four non-US countries, and neither in the US, where the four questions were How much is $100 in 2000 worth today?, Who benefits from inflation?, What really causes inflation? and Is inflation a good or bad thing?.

Is that just noise?

A gap this large should make you suspicious, so we checked it. We repeated six of the searches in the US and the UK and compared each country with its own first run:

The same US search twice shared 3.83 of 4 questions on average, so the US box is very stable. The same UK search twice shared 2.67 of 4. And the US against the UK shared 0 of 4 in both runs for all six seeds. So the difference is real: the questions in the US box are mostly different from the questions in the other four, and not because the lists are unstable. We did not test why. Whatever the cause, a US list does not describe the other markets.

The country setting does not fully localise the box

Some questions name the country: What is causing inflation in Australia? in Australia, What's causing inflation in Canada? in Canada, and What is the best bank to be with in the UK? in the UK. But others show that the box is not tightly tied to the country. For best credit card:

gb  Who has the best credit card now? | What is the best bank to be with in the UK? | Does Ambani use a credit card? | ...
au  Which credit card has the highest rating in India? | Does Ambani use a credit card? | ...
ca  Who has the best credit card now? | Does Ambani use a credit card? | What is the best premium credit card in Canada? | ...
us  What is the #1 credit card to have? | Which credit card is best overall? | What kills credit scores fastest? | ...

An Australian searcher’s box included a question about India, and Does Ambani use a credit card? appeared in the UK, Australia, Canada and India but not the US. So expect country-flavoured questions, but not only local ones.

The code

Put your key in an environment variable:

export CRAWLERDATA_API_KEY="YOUR_API_KEY"

import os
import time
from concurrent.futures import ThreadPoolExecutor

import requests

API_URL = "https://api.crawlerdata.com/v1/jobs"
HEADERS = {"Authorization": f"Bearer {os.environ['CRAWLERDATA_API_KEY']}"}


def paa(query, country="us", retries=4):
    """The People Also Ask questions Google shows for `query` in `country` (empty list if none)."""
    body = {"endpoint": "google-search", "params": {"query": query, "country": country}, "queue": "live"}
    for attempt in range(retries):
        resp = requests.post(API_URL, json=body, headers=HEADERS, timeout=250)
        if resp.status_code == 429:
            time.sleep(int(resp.headers.get("retry-after", 10)))
            continue
        resp.raise_for_status()
        job = resp.json()["job"]
        if job["status"] == "completed":
            return [q["question"] for q in resp.json()["result"]["people_also_ask"]]
        if "retry" in (job["error"] or "").lower():  # "service is busy" and "temporarily blocked" both say retry
            time.sleep(5 * (attempt + 1))
            continue
        raise RuntimeError(job["error"])
    raise RuntimeError("Search failed after retries")


def norm(question):
    return " ".join(question.lower().rstrip("?").split())


def compare_paa(seed, countries=("us", "gb", "au", "ca", "in"), workers=3):
    """Questions per country, and how many questions each pair of countries shares."""
    with ThreadPoolExecutor(workers) as pool:
        questions = dict(zip(countries, pool.map(lambda c: paa(seed, c), countries)))
    sets = {c: {norm(q) for q in qs} for c, qs in questions.items()}
    shared = {f"{a}/{b}": len(sets[a] & sets[b]) for i, a in enumerate(countries) for b in countries[i + 1:]}
    return questions, shared

The retry condition matches the word retry rather than one exact message, because we saw two different transient failures: The service is busy. Please retry shortly. and Google temporarily blocked the request. Please retry. Failed requests are not charged.

For how to learn python it printed:

us 4 ['How can I learn Python by myself?', 'Is Python easy to learn?']
gb 4 ['Is Python easy to learn for students?', 'Is there a PDF book about Python programming for beginners?']
au 4 ['Is Python easy to learn for students?', 'Is there a PDF book about Python programming for beginners?']
ca 4 ['Is Python easy to learn for students?', 'Is there a PDF book about Python programming for beginners?']
in 4 ['Is Python easy to learn for students?', 'What are the most important Python libraries?']
shared: {'us/gb': 0, 'us/au': 0, 'us/ca': 0, 'us/in': 0, 'gb/au': 2, 'gb/ca': 4, 'gb/in': 1, 'au/ca': 2, 'au/in': 1, 'ca/in': 1}

The US shared nothing with anyone, while the UK and Canada shared all four questions.

What to do with it

• Collect People Also Ask per market. One country’s list will not describe another, and the US list described the others least of all.
• Store the country with every question, as you store it with a rank.
• Expand each market on its own. Our expansion function takes a country, and the tree for each market will differ.
• Read the questions before you reuse them. Some name the country, and some belong to a neighbouring one.
• Compare with the rest of the series. Google results by country and autocomplete by country show the same kind of gap.

What it costs

Every request is 1 credit, in every country, on every queue. Starter is $1.00 per 1,000 requests and the largest pack is $0.30 per 1,000, see pricing.

Failed requests are not charged.

Next steps

You now have numbers on how far People Also Ask moves between five markets, a check that it is not noise, and a function that compares any seed across countries. The Google Search API page lists every parameter with a full example response.