The word "consultant" covers everyone from a genuine operator to someone who read three books and bought a headshot. Since the title is unregulated, you have to evaluate the work instead of the label.
What a real engagement produces
Not insight. Insight is cheap. A real engagement produces change you can point at:
- A diagnosis backed by your numbers, not industry generalities
- A decision you were avoiding, made
- A system installed and running — process, cadence, or tooling
- A measurement that shows whether it worked
- A handover so it survives without the consultant
If the deliverable is a strategy document and a goodbye, you paid for research.
The questions to ask before you sign
"What will be different in ninety days, and how will we measure it?" A vague answer here predicts a vague engagement.
"Have you run a business, or only advised them?" Both can be useful. They are not the same, and you should know which one you are buying.
"Who does the work?" In small firms, the person who sells is often not the person who shows up. Ask directly.
"What does the end look like?" Anyone who cannot describe the exit is selling a subscription.
"Tell me about an engagement that did not work." Everyone has them. People who claim otherwise are either new or not being straight with you.
Warning signs
- Deliverables measured in pages. Volume is not value.
- Frameworks before questions. If the model was chosen before your situation was understood, you are being fitted into a template.
- No numbers requested. Any serious diagnosis starts with your pipeline, margin, and cash.
- Guaranteed outcomes. Nobody controls your market, your team, or your execution.
- Scope with no owner on your side. Engagements fail on the client side more often than people admit. If nobody internally is accountable, it will drift.
What you owe the engagement
Access, honesty, and decisions. The most common reason good work produces nothing is that the owner would not make the two hard calls the diagnosis pointed to — a bad hire, a losing service line, a partner problem. No external party can make those for you.
How to scope the first one
Start small and bounded. A defined diagnostic with a clear deliverable and a decision point at the end is a far better first step than a twelve-month retainer with a stranger. If the first piece produces something real, extend it. If it does not, you learned that cheaply.
The operator difference
My bias, stated plainly: I would rather build the thing than describe it. That means the engagements I run end with a pipeline configured, a scoreboard live, a process documented, or a platform shipped — not with recommendations for someone else to implement later.
Whether you work with me or not, hold whoever you hire to that standard. Ask what will exist at the end that does not exist now.