Discussion starter: an agency question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.
- Agency
How do you evaluate a candidate’s claimed SEO results when every case study they have is under NDA?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
Evaluate how they think, not the numbers they claim. Client results under NDA cannot be verified, and headline figures are the easiest part of a case study to inflate. What cannot be faked in an interview is the ability to explain a real problem in detail: how it was diagnosed, what the evidence showed, what was tried, what did not work, and why. Combine a structured walkthrough of past work, anonymised evidence where the candidate can share it, a practical exercise using realistic data, and references from people who saw the work. Candidates who have genuinely done the work usually welcome this; candidates relying on inflated claims struggle with the detail.
What to ask about past work
- Walk me through the situation when you started: what was the site, what was wrong, how did you know?
- What data did you look at first, and what did it show?
- What did you try that did not work?
- How did you decide what to do first?
- Who implemented the changes, and what went wrong during implementation?
- How did you measure the result, and what else changed in the same period?
- What would you do differently now?
Listen for specifics: report names, the exact nature of the problem, the order of events, and honest admissions of uncertainty. Vague answers, results without method, and no mention of failures are warning signs.
Evidence that respects confidentiality
Evidence What it shows Anonymised charts with site identifiers removed That the candidate has real data and understands it Redacted audit documents or specifications Quality of thinking and communication Public writing, talks or contributions How they reason in public, where claims can be checked References from former managers or clients Independent confirmation of their role and impact A practical exercise
Give candidates a realistic scenario with data: a Search Console export showing a traffic drop, a sample of log data, and a note about a recent release. Ask them to explain what they would check, what they think happened, and what they would recommend, in an hour or two. Alternatively, ask them to review an AI-generated audit and say which findings matter and which are noise. Exercises like these show diagnostic skill and judgement far better than stories about past results.
Hypothetical example:
A candidate’s CV claims they “grew organic traffic by 400%” for a major retailer. In the interview, asked to walk through it, they describe the starting point in general terms, cannot say what the main technical issue was, and attribute the growth to “content and links”. A second candidate with a modest CV explains, in detail, how they traced a traffic drop to a faceted navigation change that multiplied crawlable URLs, the log evidence, the fix and the three-month recovery, including a mistake in the first attempt. In the practical exercise, the second candidate identifies a planted rendering problem within twenty minutes. The team hires the second candidate.
Watch for this:
Large percentage claims without a baseline are common and often misleading: 400% growth from a tiny base, or growth that coincided with a rebrand, seasonal peak or reporting change. Ask for the baseline, the period, and what else changed.
Check references properly
Ask references specific questions: what was the candidate’s role, what did they personally do, how did they handle problems, and would you work with them again. People are often more candid about role and contribution than about results. A candidate who claims to have led a project but was described by their manager as supporting it has told you something important.
Ask about the numbers they would distrust
A revealing question is to ask candidates which metrics or claims they would be sceptical of, and why. Strong practitioners can explain why average position misleads across query sets, why impressions changed in 2025 for reasons unrelated to rankings, why a single case study without a control proves little, and why AI visibility scores need careful interpretation. Candidates who accept every number at face value in an interview are likely to do the same with client data. Healthy scepticism about evidence is one of the best predictors of good judgement in this field.
Match the evaluation to the role
For technical roles, weight the diagnostic exercise heavily. For strategy and management roles, weight how the candidate explains trade-offs, prioritises and communicates with stakeholders. For content roles, ask for writing samples and an editing exercise. In every case, the question is the same: can this person reason well about search problems they have not seen before?
Evaluate process through detailed walkthroughs, accept anonymised evidence and public work, use a practical exercise with realistic data, check references on role and contribution, and treat unverifiable headline numbers as the least informative part of any candidate’s story.
Need help with this on your own site?
Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.
Add a reply Cancel reply
Replies are open to approved members. Log in or apply to join.