The phrase "AI consulting" has been used so widely in recent years that its meaning has blurred. For some firms it is a new name for selling a software subscription. For others it is a method for solving a real problem in your business. This article helps you tell the two apart, because knowing which one you are buying protects your budget and your time.
What tool selling looks like
In a tool-led sale, the conversation usually starts with the product. You are shown a dashboard, given impressive demos, and told that "this AI solves everything." The catch is that you have not even discussed which problem you are solving yet.
- The meeting focuses on product features rather than your process.
- Success is measured by usage time or license count, not by business outcome.
- Integration and training are left as "your job."
- A few months later the tool sits unused on the shelf.
The tools themselves are not bad; many are powerful. The problem is selling the tool before the problem. It is like selling medicine without asking what illness you have.
Where real consulting begins
A problem-led consultant listens first. They try to understand how your business runs, where you lose time and money, and which decisions are made too slowly. Tool selection comes at the end of that conversation, not the start.
Questions a good consultant asks
- How many hours go into this process each month, and what does that cost?
- Where do errors happen, and how long do they take to fix?
- Who does this work today, and could their time go to more valuable work?
- If we solve this, what concretely will you have six months from now?
Good consulting does not begin by telling you which tool to use; it begins by finding together which problem is worth solving.
Practical ways to tell the two apart
To work out which side the person across the table is on, watch for these signals during a meeting.
Positive signals
- They ask about your process first and leave the product for last.
- They honestly say what AI cannot solve, too.
- They propose a small, measurable first step rather than a giant transformation.
- They define success by your business goal (for example, "cut reconciliation time in half").
Warning signals
- They offer a ready solution before hearing your problem.
- They are locked into a single tool or brand.
- They describe results as exaggerated and guaranteed.
- They never mention an exit strategy, that is, the cost of leaving the tool.
So what should you do?
Before starting an AI project, prepare on your side too. Write down the three processes where you lose the most time. For each, answer the question, "What do I gain if this is solved?" Whatever consultant you work with, this list pulls the conversation to the right place: from the tool to the problem.
In the end, AI is not a goal but a tool. The right consultant shows you not the shiniest tool but the solution that creates the most difference for you. And sometimes that solution may even be a simple process fix with no AI at all. The person who can honestly say that is the right consultant.
