When should you bring in an outside consultant, and when shouldn't you?

I've been on both sides of this. I've been the consultant brought in to solve something that turned out to be unsolvable by anyone outside the organisation — and I've watched clients make real, lasting progress because they had someone external holding the thread. Neither outcome was predictable from the outside. What I've come to think, after years of doing this work, is that the question "should we bring in a consultant?" is almost never the right starting question. The better questions are more uncomfortable, more specific, and much easier to avoid asking. So this is my attempt to be honest about the conditions that make external consultancy genuinely useful — and the ones that quietly guarantee it won't be. I'm still working some of this out, so take it as a working note rather than a verdict.

the cases where outside help has a real structural advantage ¶

There are situations where the problem of being inside the organisation is the problem. You can't see the org chart as a power structure when you depend on it for your salary. You can't tell the CEO her strategy has a fatal assumption when your mortgage depends on her goodwill. An external person — one who genuinely doesn't need the relationship to survive — can say the true thing in the room. That's not a small thing. I've watched it shift decisions that were years stuck. The other structural advantage is pattern recognition across contexts. If I've seen thirty organisations navigate a product-market fit crisis, I have a library of what tends to happen next that an internal team, experiencing it for the first time, simply doesn't have yet. That library is the real product, not the slides.

when the problem is actually a political problem dressed up as a technical one ¶

This is the trap I fell into most often early in my practice. A company hires you to solve what they describe as a process problem, or a strategy problem, or a data problem. You do the work. The recommendations are sound. Nothing changes. What you didn't see — and what they possibly didn't fully see either — is that the real problem was that two senior people disagreed about the direction of the company and were using the consultancy engagement as a proxy war. No amount of good analysis resolves that. I'm not saying consultants can never help with political dynamics — sometimes naming the dynamic out loud is the intervention — but if the people commissioning the work don't actually want the problem solved, only the appearance of having tried to solve it, the engagement will fail. The tell is usually visible before you start, if you're willing to look: Who is sponsoring this? What happens to them if the recommendations are uncomfortable? Are they asking for analysis or for cover?

the honest cases where you shouldn't bring someone in ¶

You shouldn't bring in a consultant when you already know what needs to happen but lack the will to do it. I say this with some care because it can sound harsh, but I mean it practically: if your team has already diagnosed the issue, if the internal reports are sitting in a folder unread, if you've had the same conversation in three all-hands meetings — adding an external voice will not manufacture the organisational courage to act. It will cost you time and money and produce a document that joins the others in the folder. Similarly, you shouldn't bring someone in when the domain knowledge required is so specific to your organisation — your particular product, your particular customer base, your particular regulatory history — that the ramp-up time exceeds the value of the outside perspective. There are engagements where I've honestly told a prospective client: hire a fractional person from your industry, not a generalist like me. That's not a comfortable thing to say when you need the work, but it's the right thing.

the questions worth sitting with before you make the call ¶

I keep a short list I share with anyone who asks me whether they should hire external help. Not because it gives a clean answer, but because the discomfort of answering honestly usually points somewhere useful. The questions: Can you describe the specific outcome you want in one sentence, not a direction, an outcome? If the engagement produced nothing usable, who inside the organisation would be accountable for that? Is there someone internal who could do this work if we gave them the time and the permission? What would have to be true for you to act on the recommendations, even if they were difficult? That last one is the most important. If the honest answer is 'we'd need board sign-off we're unlikely to get' or 'we'd need to restructure a team we're not prepared to restructure', then the engagement's ceiling is already visible. Worth knowing before you start.

what a genuinely useful engagement looks like from the inside ¶

The best engagements I've been part of — the ones where something actually changed — had a few things in common. There was a clear, named internal owner who cared personally about the outcome, not just professionally. There was an explicit conversation at the start about what would happen with the output: who would see it, who had veto power, what the decision process looked like. And there was a willingness, from the client side, to be told something they didn't already believe. That last condition is rarer than it sounds. I've had clients who were genuinely open to being wrong about something significant. Those engagements are the ones I still think about. The others, the ones where I could feel the brief narrowing in real time toward a conclusion that had already been decided — those are the ones that taught me the most about what not to agree to next time.

a note on the consultant's own incentives (which also matter) ¶

I'd be writing an incomplete piece if I didn't say this: consultants have incentives that don't always align with your interests. An engagement that ends cleanly and quickly, with a crisp recommendation you can implement without further help, is not in my financial interest — even if it's exactly what the situation calls for. Good consulting practice means noticing when you're unconsciously scoping work larger, or introducing complexity that justifies a longer runway, or building dependency rather than capability. I try to catch myself doing this. I don't always succeed. The ask for any client is: notice if the engagement keeps expanding without a clear reason tied to your original problem. That expansion is sometimes legitimate. Sometimes it's the consulting equivalent of a restaurant that keeps refilling your bread basket so you order more wine.

so — when should you actually do it? ¶

Bring in an outside consultant when: the problem genuinely requires perspective that isn't available internally; when the internal person who could do the work is too close to the politics to do it safely; when you need speed and don't have the luxury of building the capability from scratch; or when you need a credible external voice to shift a conversation that's stalled internally (and you're prepared to act on what that voice says). Don't bring one in when: the diagnosis is done and the problem is will, not knowledge; when the brief has been written to confirm an existing conclusion; when the domain is so specific that the ramp-up cost exceeds the insight value; or when the decision-making structure doesn't actually allow for the recommendations to land anywhere. None of this is clean. Real situations are messier than any framework. But sitting with these conditions honestly, before the contract is signed, saves a lot of time — for everyone.

I don't think there's a formula here, and I'd be suspicious of anyone who offered one. The honest answer is that it depends — but it depends on specific, nameable things, not vague vibes. If you're sitting with the question right now, the most useful thing I can suggest is to answer the four questions above in writing, before you talk to anyone external. What you write will tell you more than any sales call will.