SEO interview questions
Thirteen interview questions for SEO specialist roles — role-specific, AI, behavioral, and situational — each with what to listen for, an example answer, and red flags. Pick five or six per round; a good SEO can explain why traffic moved, not just report that it did.
All 13 questions
Role-specific SEO interview questions
- 01A site you manage lost 40% of organic traffic in three weeks. Walk me through what you'd do, in order.
- 02You have budget for ten pages this quarter. How do you decide which keywords get them?
- 03Which technical SEO issues would you fix first on a mid-size site, and which would you deliberately ignore?
- 04What does a healthy internal linking structure look like, and how would you fix a weak one without a redesign?
AI questions for SEO candidates
- 05How would you make our site more likely to be cited by AI assistants like ChatGPT, Perplexity, and Google's AI Overviews?
- 06Rankings on a money keyword are stable, but clicks dropped 30% because an AI Overview now answers the query. What do you do?
- 07Where do you use AI tools in your own SEO work today, and where have you decided not to trust them?
Behavioral SEO interview questions
Situational SEO interview questions
- 11The product team wants to move the blog to a subdomain next sprint. You have two days to respond. What do you send them?
- 12You inherit a site with 5,000 thin, auto-generated pages. Keep, improve, or delete — how do you decide?
- 13Leadership asks you to guarantee a top ranking for a competitive term within 90 days. How do you respond?
Questions 1–4
Role-specific SEO interview questions
A site you manage lost 40% of organic traffic in three weeks. Walk me through what you'd do, in order.
A strong answer segments the loss by page, query, and date, rules out tracking breakage and seasonality, then separates lost rankings from lost clicks at stable rankings before proposing any fix.
What to look for
- Segments by page group, query type, device, and exact start date.
- Checks boring causes first: analytics tags, noindex accidents, quiet migrations.
- Separates lost rankings from lost clicks at stable rankings.
Example answerRed flags
Example answer
First I'd confirm the loss is real, since GA tagging breaks more often than Google does. Then I'd split the Search Console decline by page group and query type. Rankings held but clicks fell? The SERP changed — check what sits above us now. Rankings fell? Line the date up against our deploy log and known algorithm updates.
Red flags
- Proposes a fix ("build links", "refresh content") before any diagnosis.
- Never checks whether measurement itself broke.
- Blames "the algorithm" as a full explanation.
You have budget for ten pages this quarter. How do you decide which keywords get them?
A strong answer weighs intent and winnability over raw volume: who searches, whether they can buy, whether sites of similar authority already rank, and consolidates variants into one page per topic.
What to look for
- Leads with intent: who searches and whether they become customers.
- Checks winnability honestly against who ranks today.
- Consolidates: one page per topic, not per phrasing variant.
Example answerRed flags
Example answer
I'd score keywords on three things: commercial intent, evidence we can win — sites like ours already in the top five — and whether we can say something the current results don't. Volume comes last. I'd rather own ten mid-tail queries that bring buyers than chase one head term for a year.
Red flags
- Ranks purely by search volume or difficulty scores.
- Never checks the actual SERP before committing a page.
Which technical SEO issues would you fix first on a mid-size site, and which would you deliberately ignore?
A strong answer fixes crawl and index problems on money pages first, ties each fix to log or Search Console evidence, and names audit warnings it would deliberately skip.
What to look for
- Prioritizes crawl and index problems on valuable pages over cosmetic warnings.
- Names specific issues to ignore, with reasons.
- Ties fixes to logs or Search Console coverage, not severity colors.
Example answerRed flags
Example answer
First, anything blocking indexation of pages that make money: accidental noindex, canonical mistakes, broken internal links to key templates. Then crawl waste, if the site is big enough for it to matter. I'd ignore most of a typical audit: missing alt text on decorative images, 'low text-to-HTML ratio'. Fixing those moves nothing.
Red flags
- Treats the audit tool's output as a top-to-bottom to-do list.
- Can't name a single issue they would skip.
What does a healthy internal linking structure look like, and how would you fix a weak one without a redesign?
A strong answer routes internal links by page value — money pages get links from strong, topically related pages — and offers fixes that ship without a redesign: contextual links, hub pages, orphan cleanup.
What to look for
- Link placement follows page value, not just navigation.
- Cheap fixes: in-body links, hub-and-spoke clusters, orphan pages.
- Treats anchor text as information, not a stuffing opportunity.
Example answerRed flags
Example answer
Healthy means every important page is reachable in a few clicks and gets links from topically related pages with authority of their own. On a weak site I'd start with orphaned money pages, add contextual links from the strongest existing content, then build hub pages per topic cluster. All of that ships without touching the template.
Red flags
- Only talks about the navigation menu and footer.
- Suggests exact-match anchor text everywhere.
Questions 5–7
AI questions for SEO candidates
2026 · AIHow would you make our site more likely to be cited by AI assistants like ChatGPT, Perplexity, and Google's AI Overviews?
A strong answer covers three layers: crawler access for AI bots, pages that answer specific questions directly under matching headings, and third-party corroboration a model can verify across sources.
What to look for
- Knows AI crawlers need access and server-rendered text.
- Structures content as liftable direct answers near matching headings.
- Values third-party corroboration: reviews, mentions, coverage models can cross-check.
Example answerRed flags
Example answer
Three layers. Access: check the relevant AI crawlers aren't blocked and the content is server-rendered text. Structure: each page answers its core question in two self-contained sentences under the heading — that's the passage a model quotes. Reputation: models recommend what they can verify elsewhere, so reviews and third-party mentions matter as much as on-site work.
Red flags
- Claims a special file or schema type guarantees AI visibility.
- Dismisses AI search as a fad not worth engineering for.
Rankings on a money keyword are stable, but clicks dropped 30% because an AI Overview now answers the query. What do you do?
A strong answer accepts some clicks aren't coming back, competes for the citation inside the Overview with quotable structure, and shifts the keyword portfolio toward queries where users still click.
What to look for
- Doesn't promise the old click curve back.
- Competes for the citation: direct answers, extractable structure, freshness.
- Rebalances toward queries AI answers poorly: comparisons, local, transactional.
Example answerRed flags
Example answer
First I'd check who the Overview cites and what those pages do that ours doesn't — usually a cleaner direct answer near the top. I'd restructure our page to be quotable, since cited pages keep getting clicks. In parallel I'd shift the content plan toward query types where users still click through, and reset the traffic forecast with leadership honestly.
Red flags
- Treats it as a ranking problem and proposes classical on-page tweaks only.
- Hides the change from stakeholders instead of resetting expectations.
Where do you use AI tools in your own SEO work today, and where have you decided not to trust them?
A strong answer names specific tasks where AI saves hours — clustering, briefs, log analysis — plus a boundary drawn from a real failure, and a habit of verifying outputs before anything ships.
What to look for
- Concrete uses with a workflow, not a tool-name list.
- A boundary drawn from experience: something AI got wrong.
- Verifies outputs against data before shipping.
Example answerRed flags
Example answer
I use LLMs daily for keyword clustering, first-draft briefs, and turning log files into summaries — hours saved every week. I don't let them write final content unedited: one lightly-edited draft I published invented a statistic, and I stopped trusting facts I hadn't sourced. Anything with a number gets checked against the primary source before it ships.
Red flags
- Either extreme: uses AI for nothing, or publishes model output unverified.
- Can't describe a single concrete workflow, only tool names.
Questions 8–10
Behavioral SEO interview questions
Tell me about a change you shipped that hurt rankings. What happened next?
A strong answer owns a real mistake, shows the damage was caught by monitoring rather than luck, and ends with a specific process change. A candidate with no such story hasn't shipped much.
What to look for
- Owns the decision instead of blaming a colleague or Google.
- Detected the damage through monitoring, not luck.
- Ends with what changed in their process afterwards.
Example answerRed flags
Example answer
I consolidated twelve service pages into four, expecting stronger pages to rank better. Traffic dropped 25% because two deleted pages held long-tail rankings I hadn't mapped. A rank-tracking alert caught it within a week; I restored both pages. Now I never merge anything without a query-level inventory of what each page ranks for.
Red flags
- Claims nothing they shipped ever hurt performance.
- The damage was found by accident or by the client.
Describe a time a stakeholder rejected your SEO recommendation. What did you do?
A strong answer treats rejection as a prioritization problem: understand the objection, shrink the ask or reframe it in the stakeholder's terms, prove it small, and know when to drop it.
What to look for
- Understood the objection instead of repeating the pitch louder.
- Shrank scope or reframed in cost, risk, or revenue terms.
- Knows when a battle isn't worth the relationship.
Example answerRed flags
Example answer
Engineering rejected server-side rendering as a quarter-long project. Instead of escalating, I asked for one template, the product page, and we measured it: indexation dropped from days to hours and those pages gained 18% organic clicks. That number got the full migration approved the next quarter.
Red flags
- The story ends at the rejection, with no second attempt.
- Escalated over the stakeholder's head as the first move.
Describe a time you had to choose between what a guideline said and what was clearly working. What did you pick?
A strong answer shows the candidate understands why the guideline exists, sizes the risk of departing from it out loud, and makes a deliberate call they can defend either way.
What to look for
- Understands the reasoning behind the guideline, not just its text.
- Sizes the risk: what happens if this stops working?
- Made a deliberate decision and can defend it.
Example answerRed flags
Example answer
A client's competitors were buying guest posts and outranking us. I laid out both paths: match them and accept penalty risk on a domain that was the whole business, or go slower. We chose slower, and a data study earned links passively. Four months longer, and I'd make the same call — the downside was existential.
Red flags
- Follows rules blindly, or breaks them with no risk assessment.
- Pretends the tension never comes up in SEO.
Questions 11–13
Situational SEO interview questions
The product team wants to move the blog to a subdomain next sprint. You have two days to respond. What do you send them?
A strong answer sends a one-page decision document: how much traffic and revenue the blog drives, a clear recommendation, alternatives that solve the underlying problem, and a fallback plan if overruled.
What to look for
- Quantifies exposure first: traffic and revenue the blog drives.
- Clear recommendation plus a fallback if overruled.
- Asks why the move is wanted; the real problem may differ.
Example answerRed flags
Example answer
I'd send one page: the blog drives X% of organic revenue; subdomain moves historically risk ranking loss because authority doesn't transfer cleanly; here are two alternatives that solve your infrastructure problem. If it must happen, here's the redirect plan, the monitoring, and the traffic dip we should budget for.
Red flags
- Answers with a technical essay instead of a decision recommendation.
- Refuses to engage with the constraint ("just don't do it").
You inherit a site with 5,000 thin, auto-generated pages. Keep, improve, or delete — how do you decide?
A strong answer segments the 5,000 pages by traffic, impressions, and links, then treats each segment differently: improve or consolidate what earns signals, remove what earned nothing over a year.
What to look for
- Segments first: traffic, impressions without clicks, neither.
- Different treatment per segment: improve, consolidate, or remove deliberately.
- Mentions index bloat and sitewide quality as reasons to prune.
Example answerRed flags
Example answer
I'd pull Search Console and analytics for all 5,000. Pages earning traffic or impressions get improved or consolidated into stronger pages. The tail with zero signals over twelve months returns 410, with redirects only where a real equivalent exists. Then I'd watch crawl stats for a quarter — pruning at that scale usually helps the rest of the site.
Red flags
- Blanket answer: delete everything or keep everything.
- Redirects all 5,000 pages to the homepage.
Leadership asks you to guarantee a top ranking for a competitive term within 90 days. How do you respond?
A strong answer declines the guarantee plainly, explains that nobody can promise rankings, and immediately offers what can be committed: a 90-day plan, leading indicators, and a decision checkpoint.
What to look for
- Refuses plainly: anyone guaranteeing rankings is selling something.
- Offers commitments instead: deliverables, indicators, review points.
- Keeps the tone collaborative, not corrective.
Example answerRed flags
Example answer
I'd say I can't guarantee a ranking, and anyone who does is guessing with your money. What I can commit to: here's what ships in 90 days, the indicators we'll see first if it's working, and the checkpoint where we decide to double down or change course. That promise I can keep.
Red flags
- Makes the guarantee to win the room.
- Says "it depends" and offers nothing concrete in its place.
Scoring rubric
| Score | Evidence anchor |
|---|---|
| 1 | Recites definitions and tool names. No diagnostic order, no real examples, no awareness that search changed after 2022. |
| 2 | Knows classical on-page basics but jumps to fixes without diagnosis, and answers AI-era questions with buzzwords or dismissal. |
| 3 | Solid fundamentals: segments problems before acting, prioritizes sensibly, gives one concrete past example with a measurable result. |
| 4 | Strong across all four sections: evidence-driven answers, honest about failures and metric limits, credible plan for AI-answer visibility. |
| 5 | Teaches you something: original diagnostic frameworks, numbers from memory of their own cases, and a tested point of view on AI search you couldn't have gotten from a blog post. |
Frequently asked questions
Which SEO specialist interview questions work for a first screening round?
The three AI questions plus the traffic-decline diagnostic give a reliable 20-minute signal — they're hard to bluff. Kira can run this subset as a voice interview with every applicant before you spend a human hour.
How should I adapt these for an SEO manager interview?
Keep the diagnostic and AI questions, then reframe the rest around judgment at one remove: verifying an agency's report, splitting budget, coaching a specialist through a rankings drop. The seniority notes below give per-level guidance.
What are good SEO interview questions for 5 years of experience?
Use the full set, but demand specifics: real numbers from their own projects, a failure they owned, a current opinion on AI-answer visibility. Five years in, vague answers are a choice.
Do these SEO job interview questions work for agency and in-house roles alike?
Yes, with different weighting: agency roles lean on the stakeholder and reporting questions — the job is half persuasion. In-house, weight prioritization and internal linking, since the constraint is engineering time.
Should candidates prepare answers using the STAR method?
For the behavioral section, yes — situation, task, action, result keeps stories concrete. Push on the result: candidates who trail off before naming an outcome usually don't have one.
Seniority notes
- Junior (0–2 years)
- Drop the leadership scenarios and weight the role-specific section. Expect textbook diagnostic order rather than war stories, and probe whether they experiment on their own sites.
- Mid-level (2–5 years)
- The full set applies as written. Expect at least one shipped-and-backfired story and real fluency in the AI-era questions.
- Senior / 5+ years experience
- Push every answer up a level: how they'd build the process a team follows, where they'd set the AI verification policy. Ask for the numbers behind their biggest win — seniors who can't recall their own results are reciting someone else's.
- SEO manager
- Add hiring and vendor judgment: how they'd interview an agency, catch inflated reporting, and split budget between content, technical work, and links. The role-specific questions become delegation questions — what they'd do themselves versus assign.