People Also Ask: We Checked Our FAQs Against the Questions Google Shows

We pulled the People Also Ask block for six queries our own posts target, then compared the 24 questions Google showed against the 31 FAQs we had written for those same six pages. Nine of the 24 matched. Of the 18 questions that were actually about the query, exactly half had a matching FAQ on our page. On one post, the answer was zero out of four.

People Also Ask is the expanding list of related questions Google inserts into a results page. Each question opens to a short answer lifted from a page, with a link to the source. It is the cheapest read available on what a searcher wants next, and this is what it looked like when we measured it properly rather than eyeballing it.

What we measured, and how

Six queries, each one the primary term of a post published on this blog, so that every question could be checked against something we had actually written. The SERP data was read from Ahrefs' index on 3 November 2026, for the United States, taking the rows typed question in the top ten positions. Ahrefs had captured those six result pages between 17 and 30 October 2026, so the questions are up to seventeen days older than the read. Our own FAQ questions were read on 3 November from the FAQPage JSON-LD served on each live post. Hasten AI ran this analysis: it pulled the SERP rows and the FAQ questions, applied the two rules below and counted the result.

We did not query Google directly. Google's spam policies classify automated access to Search as machine-generated traffic, so a third-party index is the appropriate source for a sample like this. That choice has a cost, noted in the limitations below.

Machine-generated traffic (also called automated traffic) refers to the practice of sending automated queries to Google. This includes scraping results for rank-checking purposes or other types of automated access to Google Search conducted without express permission.
Google, Spam policies for Google web search

Two judgements had to be made by hand, and both are stated so you can disagree with them:

• On topic or not. A question counts as on topic if its answer would help someone who typed the original query. "What is a good SEO budget?" under crawl budget does not, so it is marked off.
• Matched or not. A question counts as matched if one of our own FAQ questions poses substantially the same question. Wording differences do not matter, but a shared answer is not enough. "How to check for keyword cannibalization?" matches our "How do you check for keyword cannibalization?" Our featured snippets post answers "What is a featured snippet?" inside the FAQ "What is position zero in SEO?", and that still counts as a miss, because the question it poses is a different one.

Every one of the six queries returned a block, and every block held exactly four questions

Six blocks, four questions each, with no variation at all. That consistency is worth knowing before you plan around the feature: the index recorded four questions for every one of the six, so a page is competing for one of four places.

Where the block sat varied, and it sat high every time:

Query	Position of the block	Questions	Organic results above it	SERP captured
keyword cannibalization	2	4	0	17 Oct 2026
crawl budget	3	4	1	29 Oct 2026
soft 404	3	4	1	17 Oct 2026
featured snippets	3	4	0	24 Oct 2026
canonical tag	4	4	2	29 Oct 2026
schema markup	5	4	3	30 Oct 2026

On five of the six, no more than two organic results sat above the block, and on two of them none did. An AI Overview held position one on all six.

Fifteen of the twenty-four questions begin with "what"

The distribution is lopsided, and it points at one job. Fifteen questions opened with what, five with how, two with why, and two were closed questions: "Do featured snippets still exist?" and "Can you give me an example of schema markup?"

Five of the six blocks opened with a definitional question, usually the plainest possible form of the query itself. Google's first slot tends to go to the reader who does not yet know what the term means, even when the query looks advanced. If your page opens by assuming the reader is past that, it is not answering the question Google put first.

Six of the twenty-four were not about the query at all

A quarter of the sample drifted, and the drift was not random. It followed the ambiguous word in the query:

• crawl budget returned "What is a good SEO budget?", which is about agency pricing.
• soft 404 returned "Why is error 404 so famous?" and "What does 404 mean in slang?"
• canonical tag returned "What is a canonical example?", which is about the adjective.
• featured snippets returned "What does 'snippet' mean?", in the dictionary sense.
• keyword cannibalization returned "What is an example of cannibalization?", the business term.

This matters for anyone treating a People Also Ask export as a content plan. A quarter of that list will send you to write about something your reader did not ask about. The filter is the one above: would the answer help the person who typed the query?

Nine of the twenty-four had a matching FAQ on the post that targets the query

This is the finding we did not enjoy. Across six posts we had written 31 FAQs, six on the crawl budget post and five on each of the others. Eight of those 31 did the matching, covering nine of the 24 questions, because one FAQ answers two of them. Of the 18 questions that were on topic, exactly half had no counterpart in our FAQ block. Both halves of that ratio are hand-judged, so it is a tidy number rather than a firm one: move one question across either line and it stops being half.

Query	Question Google showed	On topic	Matched by our FAQ
crawl budget	What is a crawl budget?	Yes	Yes
	What does the term "crawl budget" refer to in SEO?	Yes	Yes
	What is a good SEO budget?	No	No
	What does "crawl" mean in SEO?	Yes	Yes
soft 404	What does soft 404 mean?	Yes	Yes
	How to fix soft 404 errors?	Yes	No
	Why is error 404 so famous?	No	No
	What does 404 mean in slang?	No	No
canonical tag	What is a canonical example?	No	No
	What is a canonical URL example?	Yes	No
	What is the use of canonical?	Yes	Yes
	How to make a canonical tag?	Yes	No
featured snippets	What is a featured snippet?	Yes	No
	Do featured snippets still exist?	Yes	No
	What are examples of snippets?	Yes	No
	What does "snippet" mean?	No	No
keyword cannibalization	What is keyword cannibalization?	Yes	Yes
	What is an example of cannibalization?	No	No
	How to fix keyword cannibalisation?	Yes	Yes
	How to check for keyword cannibalization?	Yes	Yes
schema markup	What is schema markup?	Yes	No
	How do I check if my website has schema markup?	Yes	Yes
	Why is schema important in SEO?	Yes	No
	Can you give me an example of schema markup?	Yes	No

The worst result was our featured snippets post: zero of its four. The block's first question was "What is a featured snippet?", and our FAQ block never asks it. The page answers it in the body, in the opening paragraph, and poses it as an H2: "What is a featured snippet and where does it appear?" The FAQ block is the one place on that page where the question is never asked.

The pattern in the misses: 23 of our 31 FAQs answer a question Google was not asking

Look at what we wrote against what Google showed and the gap has a shape. Our FAQs skew to the follow-up: "Do soft 404s hurt my rankings?", "Can schema markup get my site penalized?", "Does Google always follow the canonical tag?" Those are good questions. They are the questions of someone who already knows what the term means.

Google's block opens with the person who does not: five of the six blocks lead with a definitional question. Four of our six posts do carry a FAQ that defines the term in the query. The two that do not, featured snippets and schema markup, are the two whose blocks opened with "What is a featured snippet?" and "What is schema markup?", and neither matched. The featured snippets post comes closest and still misses: its opening FAQ defines position zero, not the featured snippet.

The FAQ block is not a place to be interesting. It is a place to be the plainest available answer to the plainest available question.

What this does not prove

Six queries is a small sample, all of them in one subject area, all of them terms this blog targets. The question sets are not a stable object either: People Also Ask is personalised and expands as you click, so a block captured on another day, in another country, or after a user interaction will differ from this one.

Three limits matter more than sample size:

• A miss is not a failure to answer. These pages answer many of these questions in their body copy: the featured snippets post defines a featured snippet in its opening paragraph. What we measured is the FAQ block specifically, because it is the one place on the page where a question is posed and answered as a self-contained pair. Not for the schema it emits: Google stopped showing FAQ rich results on 7 May 2026, which we wrote up in October 2026.
• The sample came from a third-party index. Ahrefs captured these SERPs between 17 and 30 October, not us, and not on the day we read them.
• Both classifications are judgement calls. We stated the rules and published all 24 rows so you can apply them yourself and get a different number.

Nothing here measures whether answering a People Also Ask question wins the slot. That would need tracking across time on pages where we changed the FAQ and pages where we did not, which is the follow-up, not this.

What to do with a People Also Ask block

The useful move is small and it follows from the numbers above:

1. Pull the block for your primary term and discard the questions that drift. Expect to lose about a quarter of them.
2. Check the first question in the block against your page. If it is the definitional form of your query, make sure your page answers it in one self-contained sentence, early, in the plainest wording you can stand.
3. Compare what is left against your FAQ block, not against your article. They are different surfaces, and the FAQ block is the one that commits you to a question.
4. Write the answer at the length of the question. These are short answers, and a 200-word reply to "What is a canonical example?" is not an answer, it is a section.

If you want the broader picture of which SERP features are worth targeting and which are vanity, our guide to SERP features covers the set, and AI Overviews covers the thing now sitting above all six of these blocks. And if you are counting on FAQ markup to earn a result of its own, read what happened to FAQ rich results first.

On the evidence above, the FAQ blocks on these six posts are worth rewriting. We wrote them for the reader we wanted rather than the one Google says is arriving.