These three get compared on price. They are accountable for different things, which is what should decide it.
What each is accountable for
Consultant: a recommendation. You pay for judgment, a decision and usually a document. Genuinely valuable when the question is what to do rather than how to do it.
Professional services: their product working. Deep expertise inside one platform, delivered by the company that built it. They stop at the product boundary, which is reasonable and is also where most rollouts stall.
Forward deployed engineer: a result in your environment. Your data, your integrations, your process, your edge cases. The role exists precisely because the distance between a product and one company's reality is full of things no vendor scopes.
The three minute decision
- You do not know what to do. Consultant.
- You know, and the gap is inside the vendor's product. Professional services.
- You know, and it does not work in your company. Forward deployed engineer.
Most companies asking this question are in the third situation and shopping in the first two, because those are the options vendors offer them.
What you own at the end
This is the clearest way to tell them apart, and it is worth asking all three directly.
| What exists when they leave | |
|---|---|
| Consultant | A recommendation and a plan |
| Professional services | Their product configured |
| Forward deployed engineer | A system in production, a runbook, a named owner |
The question that exposes the difference
"What will exist when you leave, and who operates it?"
The answer separates the three faster than any capability discussion, and it also separates good providers from poor ones within each category.
Where all three fail the same way
Adoption.
None of them owns it by default. The product gets configured, the plan gets delivered, the system runs, and the team keeps working the old way.
Whoever you hire, put adoption in scope explicitly: who trains, who follows up in week three, and how you will know people are actually using it. Otherwise you will be having this conversation again in six months, about a different vendor.
A note on sequencing
Consultant then engineer is a reasonable order when the strategy is genuinely unclear. The risk is paying twice for discovery, so require the consultant's output to be concrete enough to build from: specific systems, specific rules, specific definitions of done.
If the recommendation cannot be handed to a builder without a second discovery phase, it was not finished.