Hand an architect a problem and you get drawings. Beautiful ones — elevations, a materials list, a plan that accounts for the afternoon light. Hand a builder the same problem and, eventually, you get a house. Both are doing their job well. But if what you needed was somewhere to live, only one of them got you there. The drawings are not the house, however good they are.
A consultant and a builder are paid for opposite things, and most people don't notice until the money is spent. A consultant is paid to think — the deliverable is understanding: a recommendation you can act on, a decision you can defend. A builder is paid for the thing existing — the deliverable is a system that runs whether or not anyone is thinking about it. Both can be excellent. They aren't interchangeable, and the trouble starts when you buy one expecting the other.
The calendar and the commit
Watch what each one does when the work gets hard.
A consultant's instinct, faced with uncertainty, is to schedule. Another workshop, or a deeper discovery phase to align the stakeholders first. This isn't laziness — it's the trade. Thinking is the product, and there's always more thinking to sell. The calendar fills up. The thing still doesn't exist.
A builder's instinct, faced with the same uncertainty, is to make the smallest real version and see what breaks. Not because builders are braver, but because nothing they're paid for happens until something runs. A builder with an empty calendar and nothing shipped has earned nothing. A consultant with a full calendar and nothing shipped has had a good month.
You can't schedule your way to a working system. At some point the meetings have to stop and someone has to build the thing.
How to tell which one you hired
The signals are simple, and you can read them in the first month.
- Ask what will exist at the end that doesn't exist now. A builder names a system. A consultant names a document, a framework, or a "clear path forward."
- Watch where the time goes — toward writing code and shipping it, or toward more meetings about what to do.
- Count the things in production. Three months in, has anything started running, or has the deck just gotten longer?
None of this makes consultants the villains. There are real problems where the deliverable genuinely is a decision — should we enter this market, should we buy that company. For those you want a clear thinker and no code at all. The error isn't hiring a consultant. It's hiring one when what you actually needed was the thing to exist.
The honest version
We're builders, so read this with that in mind. But the test holds whoever is selling: if you need a working system, the person you want is the one whose income depends on it running, not on the next session being booked. Inverted economics, inverted incentives. Pay for thinking and you'll get more thinking. Pay for the thing to exist, and — if you've chosen well — the thing will exist.
For the technical reader
The inversion is structural, not a matter of character, and it shows up in how the two kinds of firm are staffed and paid.
An advisory firm is built around utilisation: billable hours or monthly retainers, staffed with analysts and strategists whose output is analysis. That's a different cost structure, and a different hire, from the engineers who put systems into production. It isn't that advisory firms won't build — it's that they aren't assembled to. The people who can describe the system that should exist and the people who can make it exist are usually different people, at different rates, in different firms.
The incentive follows the staffing. Hourly and retainer models are paid for elapsed time, so finishing ends the revenue and the rational move is to extend — another phase, another refinement of the strategy. Outcome and fixed-scope models are paid for a defined result, so finishing is the win and speed helps the supplier instead of costing them.
So the buying rule is mechanical. When the deliverable is genuinely a decision, buy the thinker. When the deliverable is a working system, buy whoever's paid only when it runs — and verify it the one honest way, by asking what's in production today that wasn't there before. If the answer after a quarter is a slide deck, you bought the wrong instrument for the job. Not a bad firm. The wrong instrument.