The short version
- Ask what they've actually done, ask what they can show you before you pay, and ask what they're wrong for. In that order, out loud, on a free intro call. Nothing is standing behind these people, so work they can describe is the only evidence there is.
- This is the worse instrument and I'd rather say so than pretend. A cheap standardized check is uniform, verifiable, and indifferent to whether you know anyone. Your personal judgment is expensive, inconsistent, and exactly the surface where bias operates. You're running it here because the alternative doesn't exist, not because it's better.
- SEO is the hardest of the three. Results are slow and confounded by everything else the business does, so what an honest SEO mentor shows you is sites they've moved plus an account of what else changed at the same time. Clean attribution for a ranking gain tells you about their standards, not their results.
- Mindset coaching is the one field here with real certification bodies, and they certify the coach's process rather than any client's outcome. It also shades into territory where the right professional is a licensed one, and a coach who won't name where their scope ends has told you something.
- Software development is the easy case, and that's why it's in here. Demonstrable work is abundant, so it shows you what "show me" looks like when the field cooperates.
Why these three fields are in one guide
What do SEO, mindset coaching and software development have in common? No credential in any of them tells you whether the person can do the job you're hiring them for. In two of the three there's no issuer at all. In the third there is one, and it certifies training hours rather than a client's result. Different causes, same consequence: the verification lands on you.
One of these three doesn't obviously belong, so I'd rather explain the reasoning than hide the seam.
The original version of this group was "fields with no credential." That's true of SEO and software development. Nobody licenses either: no board, no register, no exam anybody checks, and no mechanism by which a bad practitioner loses the right to call themselves one. Mindset coaching is different. It has certification bodies, some long-established, with published standards and real requirements.
So I corrected the shared property rather than dropping the page. These three fields aren't grouped by the absence of a credential. They're grouped because no available credential tells you whether this person can do what you're hiring them for. For SEO and software development that's because nobody issues one. For mindset coaching it's because the one that exists measures training and practice hours, which is a different question from whether your situation improved.
That makes mindset coaching the most useful member of the group rather than the odd one out. It's the case where a credential exists, is honestly run, and still doesn't answer your question. Without it you could tell yourself the problem is missing paperwork. The coaching case shows you the paperwork wouldn't have solved it.
Two boundaries before we start. If you're still deciding whether you want a coach or a mentor at all, that's a different question, answered properly in coaching vs. mentoring. This guide assumes you've decided you want a person and you're working out which one. And the three questions below aren't specific to these fields.
What a credential would do better than you can
Is a credential really worse than judging the person yourself? No. Pretending otherwise would be convenient for me and it wouldn't be true. A standardized check is cheap, uniform, and blind to whether you know anyone. Personal verification costs hours, varies with your mood and your reference points, and is the precise surface where bias operates. In these three fields you're stuck with it anyway.
Think about what a credential is as an instrument. One syllabus, one pass mark, one price, applied identically to a stranger and to someone whose father plays golf with the hiring manager. It doesn't care about your accent, your university, whose logo is on your resume, or whether you can get a warm introduction. That is a genuinely valuable property and it's the one thing personal judgment can never replicate. People defend credentials in fields that have them for a good reason: a thin, uniform check is the fairest instrument anybody has invented for evaluating strangers at scale.
Which means the credential-free field is worse for the outsider, not better. If you already know three working SEOs you can ask, or you can read a codebase well enough to judge one, your personal check is far better than the one available to a reader with neither. Removing the credential doesn't democratize the judgment. It hands the judgment to whoever already has access.
I'd rather say what that means for you than leave it as an observation. If you have no network in the field, the method below is genuinely weaker in your hands than in the hands of someone who does, and no amount of good questions closes that gap entirely. Two things narrow it. Ask the questions across three people rather than one, because you are trying to calibrate an unfamiliar field and comparison does work that expertise would otherwise do. And treat the first paid month as part of the evaluation rather than the start of the relationship, with something specific you expect to be different by the end of it. Someone with three working SEOs on speed dial can skip both. You can't, and pretending otherwise would be the same trick as a listicle.
There's a bias in my own advice too. "Show me your work" quietly favors people who are permitted to show work. Agency contracts, client confidentiality and in-house teams that never publish anything all produce excellent practitioners with thin public records. Someone who ran SEO inside one company for ten years may have nothing you can look at, while someone who spent those years publishing about SEO has a portfolio made mostly of publishing. Run this instrument lazily and it penalizes exactly the wrong people.
So here's the correction, and it's the most important sentence in this section. The test isn't whether there's a public portfolio. It's whether they can describe the work under questioning. A person who genuinely ran the thing can tell you what they tried first, what didn't work, what they'd do differently and what they still aren't sure about, whether or not they're allowed to name the client. That's much harder to fake than a case study page, and it doesn't punish anyone for having signed an NDA.
None of this is a celebration of credential-free fields. It's a description of a constraint. You're being asked to run the expensive, unreliable check because the cheap, reliable one was never built.
One correction to that, because it's the obvious objection and it's a fair one. The three questions below are themselves a standardized instrument: fixed, cheap, asked in the same order of everybody. So it isn't quite true that no structured check exists for these fields. What's missing is not the instrument, it's the institution. Nobody issues it, nobody audits whether you ran it, and nobody stops a person practicing if they fail it. That distinction matters for how much weight you put on your own result. A structured check beats an unstructured impression, which is the entire reason for running these in a fixed order rather than going on how the call felt. It just doesn't carry anyone's authority but yours.
What they've actually done, and how recently
What should you ask first? Ask what they've done and when, not how long they've been doing it. Ten years is a duration, not evidence. "I moved this site, in this window, and here's what else was changing at the same time" is evidence. Recency matters most where the ground has shifted recently, and since nobody enforces a recency bar in any of these fields, you have to.
SEO. There's no gatekeeper. Results arrive slowly, often over quarters. And they're confounded by everything else the business is doing: a redesign, a funding round, a competitor's outage, a product launch, a pricing change, a seasonal swing. The causal chain from a specific decision to a specific outcome is genuinely hard to establish, so the field rewards confident storytelling over careful attribution, and the loudest voices are often the people best at marketing themselves. That's a real skill and a completely different one from doing the work.
So what an SEO mentor can honestly show you is sites they've moved plus an account of what else changed at the same time. Ask directly: what else was going on? Someone who has lived through a ranking recovery answers immediately, usually with a list, because they spent the whole quarter arguing about attribution with somebody. A clean causal story about a messy outcome isn't a stronger claim. It's a lower standard. Then ask what they've changed their mind about in the last two years. In a field this unstable, an answer built entirely from old work is a recency problem you have to catch yourself.
Mindset coaching. Here the question has a partial answer on paper. The International Coaching Federation's entry-level credential asks for 60+ hours of education and 100+ hours of coaching experience, plus mentor coaching, a performance evaluation and an exam. The top level asks for 200+ hours of education and 2,500+ hours of coaching experience. Those are levels of coaching practice, not certifications in any particular subject.
Read the list again and notice what's being measured. Hours of training. Hours of coaching delivered. A session assessed against a competency framework. Those are meaningful, and it's more structure than SEO offers anybody. What they aren't is a client outcome. A credential built on practice hours tells you the person has done the reps and been observed doing them. It's silent on whether the people in those sessions got anywhere.
That's not an accusation, it's what the instrument measures. So your question stands: who have you worked with whose situation looked like mine, and what happened? A coach with a real practice has an answer, usually a careful one, often with the parts that didn't work included.
Software development. The easy case, here deliberately so you can see what a good answer sounds like when the field cooperates. Listen for "shipped and maintained," not "worked on." What did you own, who used it, what broke in production, what did you do about it? Maintenance is the tell, because anyone can start something. The person who ran a service for two years after launch has a completely different set of stories from the person who built it and moved on, and you can hear the difference in about ninety seconds.
What you can see before you pay
What can you actually check for free? Name the artifact. A profile detail you can verify, a public body of work, reviews from people whose problem resembles yours, and a free intro call where you ask one question you prepared in advance. In software development the artifact is abundant. In SEO it's scarce and confounded. In coaching it's mostly the call itself, which is why the call has to be run properly.
Start with software development, because it's the benchmark. Software development mentors can point at public repositories, commit history over years, a package other people depend on, a technical write-up that survived comments from strangers. You don't need to evaluate all of it. Commit history tells you about persistence. A library with external users tells you somebody trusted the work enough to depend on it.
There's a version of question two here that costs nothing and works better than reading their profile: instead of looking at their work, have them look at yours. Bring a piece of code to the free intro call and ask what they'd want to know about it before reviewing it properly. Watch what comes back. Someone who has genuinely maintained software asks about the thing you didn't consider, not the thing you got wrong. Anyone can find the bug. The person worth paying asks what happens when the data gets ten times bigger, or who's on call when this fails at 3am. The question they ask is the artifact, and you get it before any money changes hands, which is the whole point of this section.
SEO gives you much less, so ask for less and ask precisely. One site, one window, one account of the confounders. If they can't share client work, and plenty of good practitioners genuinely can't, ask them to walk you through a decision instead: a time they recommended not doing something, and why. Their own site is evidence, but weak evidence, because it's the one field where marketing yourself and doing the work are the same activity.
Mindset coaching hands you three checkable things. The credential claim is verifiable: ask which body, which level, and whether it's current, then check the body's own directory rather than taking the profile's word for it. Reviews are useful if you filter them to people whose situation resembles yours instead of reading the average. And the intro call is the main artifact, so run it properly.
Across all three, one rule does most of the work. Go into the call with one specific thing from your own situation, and see whether they ask you about it. Getting asked a question you have to stop and think about is the best signal available before money changes hands.
The tells that should stop you
What's the clearest warning sign? A result with no denominator. Anybody who has actually moved a site, shipped a service or coached clients for a living knows what else was happening at the same time, and says so without being asked. Clean attribution offered up unprompted isn't a stronger claim than a messy one. It's a weaker standard, and it's the thing here you can test with a single question.
These are behaviors, not personality types, and every one has a check attached. That matters, because a "tell" you can't test is just a prejudice with better branding.
- The result with no denominator. Traffic tripled, revenue doubled, the client got promoted, and nothing else in the world changed that quarter. Check: "What else was going on at the same time?" A practitioner answers with a list. Watch what happens when you ask it twice.
- Duration standing in for recency. Twelve years in the field, described entirely in the past tense. Check: "What have you shipped in the last twelve months, and what have you changed your mind about since?" The second half is the harder one to fake.
- The unscoped yes. Everything you described is something they can help with. Check: "What do you turn down?" Someone running a real practice has said no to people and remembers why. Someone who hasn't will generalize, and you'll hear it.
- Audience size doing the work of evidence. A profile whose main credential is reach. Check: ask what the largest thing they've been accountable for was, and who else was on that team. Reach is produced by publishing; the check asks for something publishing cannot generate. Someone whose audience grew out of the work answers immediately and usually names colleagues. Someone whose audience is the work goes vague on the second half.
- No named boundary, in coaching specifically. A coach who won't say where their scope ends. Check: "What's outside what you do, and who do you refer people to?" This isn't a rude question, and the field's own standards assume it gets asked. ICF's Code of Ethics requires its professionals to disclose when they're acting in a role other than coach, and includes a standard about exploring "a different coach, professional, or resource" when the value of the relationship shifts. A coach who can't answer is out of step with their own field, not just with me.
- The intro call that's a sales call. Check: did they ask you anything you had to think about? If not, you already know how the paid hour goes.
One thing that is not a tell, because I'd rather you didn't misuse the list above. A thin public portfolio is not a warning sign. NDAs, client confidentiality and in-house work all produce it, and it hits the deeply experienced hardest. What matters is whether they can describe the work under questioning. Someone who ran the thing can talk you through the decisions without naming a single client. Someone who didn't gets vague at exactly the point where the detail should be getting denser.
What each of them is wrong for
When should you not hire one of these mentors at all? When the problem you have isn't the problem they solve. An SEO mentor can't fix a business with nothing worth ranking for. A mindset coach can't do a licensed professional's job. A senior engineer can't give you reps you haven't done yet. Every genuinely good mentor is the wrong choice for somebody, and a profile that implies otherwise is itself the tell.
An SEO mentor is wrong for you if the real problem is that there's no demand for what you sell, or nothing on the site anybody would want to read. That's a product and positioning problem wearing an SEO costume, and mentoring on the costume won't touch it. They're also wrong if you want a guarantee, because the field can't produce one, and anyone offering one is telling you what they think you want to hear.
A mindset coach is wrong for you if what you're carrying needs a licensed professional. I'll say it plainly, because the field is fuzzy about it and you shouldn't have to work it out from context. Coaching is not therapy. Coaches are not licensed clinicians and coaching credentials are not clinical licenses. If what you'd be bringing to the sessions is something you'd raise with a doctor, that's a different professional, and no amount of goal-setting structure changes it. A good coach will tell you so themselves, unprompted, which is why the boundary question above is worth asking. They're also wrong if what you actually want is domain expertise. A mindset coach isn't going to review your pull request.
A software development mentor is wrong for you if you haven't done any work yet. The value here is somebody experienced looking at real decisions you made and telling you where the reasoning was wrong. With nothing to look at, the session becomes a lecture, and lectures are cheaper elsewhere. Build the thing badly first, then bring it. They're also wrong if you want an answer rather than a critique, because the useful version of this relationship is uncomfortable about a third of the time.
And any of them is wrong for you if the intro call didn't work. Walking away after a free call is the cheapest decision in this whole process, and it's the one most people skip because declining feels rude. It isn't. Nobody good is offended by it.
Which leaves me one honest version of a pitch. We hand-screen every mentor before they're listed, and I'd still rather you didn't take my word for it, because "we check them so you don't have to" is exactly the unverifiable claim this guide has been teaching you to distrust. The screen gets you a shorter list. It doesn't get you a decision. What you can check yourself is the profile, the reviews on it from people with your problem, and a free intro call where you ask the three questions above and listen to how the answers land. Start with an SEO mentor, a mindset coach or a software development mentor, then run the check on the two or three you shortlist. The check is the product. We just make the list shorter.
Frequently asked questions
How do you vet an SEO mentor?
Ask for one site they moved, the window it happened in, and what else changed at the same time. The last part is the real question. SEO results are slow and confounded by redesigns, launches, competitor activity and seasonality, so anyone who has done it can list the confounders immediately. Clean attribution offered without prompting tells you about their standards rather than their results.
Does an ICF credential mean a mindset coach is good?
It means something narrower than most buyers assume. ICF's entry-level credential requires 60+ hours of education and 100+ hours of coaching experience, plus mentor coaching, a performance evaluation and an exam. That's real structure, and more than uncredentialed fields offer. What it measures is the coach's training and practice hours, not whether their clients got anywhere. Check the credential is current, then ask about outcomes separately.
What should I ask on a free intro call?
Bring one specific thing from your own situation and see whether they ask you about it. Beyond that, three questions do most of the work: what have you done that resembles this, what can I look at before I pay, and what do you turn down? The third is the most revealing. Someone running a real practice has turned people away and remembers why.
Should I choose a mentor with a large following?
Audience size measures how well someone markets themselves, which is a genuine skill and a different one from the work you're buying. In SEO especially the two get confused, because marketing yourself and doing SEO look identical from outside. Treat reach as neutral, then run the check you'd run on anyone: what did you do for one client whose situation looked like mine, and what happened?
When should I see a licensed professional instead of a mindset coach?
When what you'd be bringing to the sessions is something you'd raise with a doctor. Coaching is not therapy, coaches are not licensed clinicians, and coaching credentials are levels of coaching practice rather than clinical licenses. That's a scope difference, not a quality judgment. A coach working to a professional standard names the boundary themselves and tells you who they refer to, which is why it's worth asking on the first call.
How do I check a software development mentor if their work is private?
Ask them to describe it instead of show it. NDAs and in-house work produce thin public portfolios for some of the most experienced people in the field, so a missing repository list isn't a warning sign. Ask what they owned, what broke in production, and what they'd build differently now. Someone who ran the thing gets more specific under questioning. Someone who didn't gets vaguer.
