A five-stage loop for a senior organic search role at a company of any size usually contains one interview with somebody who knows what a canonical tag does. The other four are a recruiter screen, a hiring manager who owns growth, a product manager, and an engineering manager whose cooperation you will need for every technical fix you ever ship.
That ratio is the whole problem. Candidates prepare for the one interview and lose the other three.
The screenshot problem
Most SEO people arrive with a deck of line graphs going up and to the right. Traffic tripled. Non-brand clicks up 240% year on year. Thirty thousand new keywords in the top ten.
A hiring manager who has run this function before will ask one question about that chart, and it is always the same question: how do you know it was you? Google shipped core updates during your tenure. The company also launched in three new countries, doubled the content team and ran a brand campaign. Your graph is consistent with your work mattering and consistent with it not mattering at all.
Have an answer ready that does not depend on the graph. Holdout groups on templated pages, a staged rollout across a page type, before-and-after on a set of URLs you touched against a matched set you did not. Any of those beat a timeline annotation. If you have never run a test like that, say so and describe how you would, because the honest version of that answer still lands better than confidence about a correlation. The same reasoning that makes a traffic drop diagnosis credible is what makes a traffic gain claim credible, and interviewers in product organisations are trained on exactly this distinction.
What the PM and the engineering manager are grading
Neither of them is checking your technical knowledge. They are checking whether working with you will cost them sprint capacity for things they cannot justify to their own leadership.
So the useful preparation is one story about a technical change you actually got shipped. Not the audit. The ship. Which engineer, what the ticket said, what you cut from the original scope to get it into a sprint, how you sized the upside before anyone had written code, and what it did afterwards. A candidate who says “I handed over a technical audit with 140 issues and the dev team never got to it” has described a failure and does not always realise it.
Pick the smallest possible version of the story. One template, one fix, a number at the end. Engineering managers trust specificity and distrust roadmaps.
The four answers worth rehearsing out loud
| What they ask | What they are testing | Where candidates lose it |
|---|---|---|
| ”Where would you start here?” | Prioritisation under a real constraint | Listing everything instead of choosing one thing and naming what you would skip |
| ”How would you measure it?” | Causal reasoning | Reporting clicks and calling it attribution |
| ”Tell me about something that did not work” | Whether you learn | A disguised success story |
| ”What would you need from us?” | Whether you can be managed | Asking for nothing, or asking for everything |
Worth knowing that the better hiring managers are actively designing around rehearsed answers now, which is why the questions you get may be built on the company’s own data rather than anything generic. We cover that from the other side of the table in how to interview SEO candidates.
The fourth one gets underrated. Say two specific things: access to the deployment process in whatever form it takes, and a named person in engineering with recurring time. Vague answers here read as someone who has only ever worked where SEO was a reporting function.
Where a paid mock interview earns its fee
Reading those four answers is not the same as saying them to a stranger who interrupts. The gap between a well-written answer and a well-delivered one is roughly the gap between a 40-minute interview going well and going badly, and you cannot close it alone in a notebook.
Prepfully is the marketplace I would point someone at for this, with one caveat first: it does not have an SEO category. Its coach roster is built around the roles the rest of your loop belongs to, which is PM, TPM, engineering manager, software engineer, data science. Roughly two thousand coaches across more than 1,600 companies, 20,000 sessions completed, 4.85 average rating, and sessions priced between about $230 and $450 depending on seniority.
Used literally, that is the wrong product for an SEO manager. Used properly, it is close to the only product that does the thing that matters, because you can book the person who sits in the chair you are trying to impress. A Google L7 product manager running a mock against your prioritisation answer will tear it apart on structure, which is precisely the skill the search-adjacent interviewers in your loop are grading. The same goes for a data science coach and your measurement story. Their question bank is also worth half an hour of reading before you pay for anything, since it is filtered by role and company and tells you what was actually asked where.
Two cases where I would skip it. If the company is fifteen people and the founder is interviewing you, there is no rubric to rehearse against and your money is better spent on the plan you bring. And if you are moving into a role where search is the whole job rather than one input into growth, a mock with a PM will polish delivery while missing the content of the interview entirely. Then the better move is the unpaid one: ask a former colleague who has hired for the role to grill you for an hour.
Bring a plan for their site, not a portfolio of someone else’s
The strongest thing a candidate can do costs nothing and almost nobody does it. Spend two hours on the company’s own site before the hiring manager stage and turn up with one page.
Three things on it. The single largest opportunity you can see from the outside, with a rough size and the assumption behind that number. One technical problem you can verify from a crawl of fifty URLs. One thing you cannot assess without internal data, and the exact question you would ask on day one to assess it. That last line does more work than the other two, because it proves you know the difference between what is visible externally and what is not, and it stops the document reading like an unsolicited pitch.
Keep it to a page. Our audit template is more than you need here; the interview version is deliberately thinner, and the point is not to be right. It is to give the hiring manager forty minutes of talking about their actual site with somebody who prepared.
One more thing on levelling. Decide your number before the recruiter screen and know which band it sits in at that company, because the first conversation often sets the band and the band sets everything else. Our pricing benchmarks are agency-side rather than salary data, but the ratio between them is a useful sanity check on what an in-house team is worth paying for.
