AI · 2026-08-04 · 11 min read · by Juan Carlos Zuloaga
Most AI consulting proposals sell a workshop. You end up with slides, a maturity score, and no change in how the company runs.
That is not consulting. That is theatre.
For a company between 50 and 500 people, useful AI consulting has one test: after 90 days, is there a process running differently, with a number attached to it?
This article explains what real AI consulting services include, how we scope them at Inspiral Growth, and how to tell a delivery partner from a slide factory.
Strip away the language and there are four things a consultant can sell you.
1. Diagnosis. Where does work pile up? Which decisions are slow? Which data already exists but is never used?
2. Prioritisation. Of twenty possible AI use cases, which three are worth doing this quarter based on effort, risk, and value.
3. Build. Someone actually ships the automation, the assistant, the pipeline, or the rebuilt page.
4. Adoption. The team uses it after the consultant leaves.
Most firms sell 1 and 2. The value sits in 3 and 4. When you evaluate a proposal, count how many days are allocated to building versus analysing. If build is under half, you are buying a report.
Small companies rarely have enough process to automate. Enterprises spend two years on procurement before anything ships.
A 50–500 person company has the useful middle: real volume, real repetition, real data — and a leadership team that can still approve a change in one meeting.
That combination means an AI engagement can move from idea to production in weeks. It is also why our engagements are structured around a decision owner, not a committee. One person who can say yes.
Sprint (2–4 weeks). One process, one automation, one number. Good for proving the approach before committing budget.
Quarter (12 weeks). A cluster of related use cases plus the internal capability to keep going — templates, prompts, documentation, and one trained internal owner.
Sparring retainer. A recurring session where leadership brings live decisions and gets an outside perspective. This is closer to our business coaching and sparring work than to classic consulting, and it suits companies that already have technical capacity but lack direction.
Most clients start with a sprint. It is the cheapest way to find out whether the working relationship is any good.
Before proposing anything, we map four things.
- Volume. What happens more than fifty times a month?
- Variance. Where do two employees do the same task differently?
- Latency. Which handoffs take days when they could take minutes?
- Evidence. What is measured today, and how badly?
The answers usually point at the same handful of places: inbound handling, quoting, scheduling, reporting, and content production. Those are boring. Boring is where the money is.
In our Clean Clean Schoonmaak partnership, the constraint was not marketing. It was that inbound requests, scheduling, and follow-up all lived in someone's head and inbox.
The work started with the least glamorous step — writing down the actual flow — and then rebuilding the front door of the business so requests arrived structured instead of as free text. Structured input is what makes automation possible at all.
The pattern repeats across our case studies: fix the input, then automate the middle, then measure the output. Skipping to step two is the most common way AI projects stall.
- What will exist in production at the end, and who owns it?
- Which of your people will actually build it?
- What happens to the work if we do not renew?
- Show me something you shipped that is still running a year later.
- What did you recommend against building, and why?
The last question matters most. A partner who has never talked a client out of an AI project has not been paying attention to cost.
Pick one process. Measure it for two weeks — honestly, with a stopwatch if needed. Then change one thing and measure again.
You do not need a consultant for that first loop. You need one for the second, when the easy wins are gone and the remaining problems touch data, systems, and people at the same time.
If you want a second opinion on which process to start with, book a discovery call: https://calendly.com/inspiralgrowth/discovery-call