Diagnose your search problem before you hire
TL;DR
When search underperforms, the reflex is to go shopping: a consultant, an agency, a tool, more budget. That's shopping before you've diagnosed. Here are the questions you can answer from your own account first, so the right help falls out of the problem instead of the search box.
When search stops delivering, the reflex is to go shopping. A consultant, an agency, a new tool, or just more budget. The month's numbers are flat, something has to change, and hiring feels like the change.
The trouble is you're shopping before you've diagnosed. You've decided the answer is "someone" or "something" without settling what's wrong, and the market is happy to sell you a solution to a problem you haven't named. The diagnosis is the part nobody sells you, because it's the part you're meant to do yourself. It's also the part that decides whether any of the shopping is worth doing.
The label picks who you meet
Start from a label rather than a problem and the label does the choosing for you. Search "SEO consultant" and you meet SEO consultants. Search "fractional CMO" and you meet fractional CMOs. Each will tell you, honestly enough, that your situation is the kind their discipline handles, because that's the situation they see all day.
So the person you end up hiring is shaped by the term you typed, and the term reflects a guess about the answer rather than a read on the problem. The founder's own writing makes this point about the two big umbrella labels: performance marketing and search marketing describe who does the work rather than what's wrong with your account, and picking one first tends to route you to whoever optimised hardest for that word. Write the problem down first and the discipline usually names itself.
The questions to answer before you brief anyone
Defining the problem sounds abstract until you turn it into questions you can answer from your own account. Four of them do most of the work.
Can you trust your own numbers? This is the least glamorous question and the most important one. If conversions are miscounted, double-counted, or firing on the wrong page, every decision downstream is built on sand, and a strategist working from those numbers is working blind. Two independent practitioner audits put a number on how common this is. Sarah Stemen, auditing underperforming Google Ads accounts, found 80% of "failing" accounts traced back to broken conversion tracking or targeting problems rather than a fundamentally broken channel (Stemen, 2026). GrowthSpree, reviewing more than 300 B2B SaaS accounts and over $60m in managed spend, found tracking broken in at least one significant way in 70 to 80% of them (GrowthSpree, 2026). Both are practitioner observations rather than controlled studies, and both land in the same range. If your account is in it, the problem isn't strategy yet. It's measurement, and no amount of senior strategy time fixes a number you can't trust.
Is this a demand problem or a visibility problem? Are people searching for what you sell, and if they are, are you showing up for them? These are different problems that feel identical from the inside, where all you notice is that the phone stays quiet. If nobody's searching, more budget buys you more of nothing and the work is about creating or capturing demand elsewhere. If they are searching and you're absent, the work is visibility, and it's a much more tractable brief.
Which surface is the problem on? A searcher sees one page with paid results, organic results, and increasingly an AI-generated answer above both, then makes one decision about where to click. Your org chart splits that single moment into two or three separate jobs with separate owners. Working out whether you're losing the click on the ad, the listing, or the answer at the top tells you which surface needs the help, and stops you paying for work on the one that was already fine.
Is it a fix or a discipline? Some search problems are a one-off repair. A broken tag, a misconfigured feed, or a page that loads slowly enough to hit both your organic rank and your ad costs at once. That last one shows why the diagnosis pays off: Google's ranking systems use Core Web Vitals, and landing page experience is one of the three inputs to Quality Score, which helps set what you pay per click. One slow page runs up two bills, and a single fix clears both. Others are ongoing management that never finishes. The two need completely different help, and conflating them is how a business ends up on a twelve-month retainer for a job that needed a fortnight of a specialist's time.
Why the market skips this step
None of these questions is hard to ask. The reason they get skipped is that problems are harder to sell against than labels.
A label rewards whoever ranked for it. A problem is specific to you and can't be gamed in advance, which makes it inconvenient for anyone whose business depends on you arriving pre-sold on the solution. The search platform's own AI won't suggest you spend less and lean on organic, because its revenue sits on the spend. A tool wants you to conclude you need the tool. A retainer wants to be a retainer. Everyone selling has a preferred diagnosis, and it tends to match what they sell. The founder's guide to finding a fractional CMO or search consultant makes the same observation about the marketplaces that match buyers to help: the differences between them come down to who does the filtering, what the filter costs, and whose interests set the sort order.
When you can name it, the help names itself
Run the diagnosis and the shape of the help usually falls out of it.
If it's a measurement problem, you need a specific technical fix, and the question becomes how to vet someone who can do it well. The founder's guide to choosing an SEO consultant is built for exactly that: the pre-call checks you can run from your desk, the red and green flags, and the one signal nobody can buy, which is published work you can verify yourself. If it's a "we've outgrown doing this in-house" problem, the fractional and consultant routes are the map, and the same discipline applies: check the work, not the label on the website. If the honest answer is "we can't tell which discipline this is", that's the signal to write the problem down before briefing anyone, because the writing is what makes the discipline obvious.
The diagnosis is also work you can do before hiring for it. Reading your own search data for what's broken, which numbers to trust, where demand sits, which surface is leaking, is reasoning you can do yourself, or with a tool built to reason through it, which is the premise Addy starts from. The definition still belongs to you either way. The point is to hold it before you spend against it.
When the diagnosis points away from search
Sometimes the honest answer is that search isn't the problem at all.
People are finding you, clicking through, and not buying. That points at the offer, the pricing, or the page they land on, and none of those is something a search hire can fix. It's the least comfortable outcome of the diagnosis and often the most valuable, because it's the one the shopping reflex skips entirely. You can't buy your way past it with a better consultant.
So the question to sit with before you brief anyone is a simple one. Could you describe the problem to a stranger in two sentences without naming a discipline? If you can, you already know roughly who to call. If you can't, that's the work to do first, and it's cheaper than any of the help you were about to go and buy.