Hiring an AI consultant for a small business: ten questions to ask every provider, including us
If you search for an AI consultant for small business, you'll find lots of offers that sound alike: workshops, strategies, potential analyses, implementation. From the outside, it's hard to tell which of them will leave a running process in your business at the end. We don't compare providers here. We give you ten questions you can ask any of them before you sign. And because we're one of those providers ourselves, each question also comes with our own answer. You can hold us to it.
Why questions help more than a comparison
A comparison of providers goes out of date quickly, and it says little about whether someone is a fit for your business. Questions, on the other hand, you can ask in every first conversation, whether it's with an AI consulting firm or an AI implementation consultant. Good answers are concrete: a process, a yardstick, a name on your team, a report. If an answer gets stuck on what the AI is capable of, keep asking until you know who checks what.
The ten questions follow a logic we apply ourselves: first the process, then the control, then cost, risk and proof. Our answers are short. Where there's more detail, we link to the article that covers it.
Questions about the process
1. Which process comes first, and why that one?
Ask how the first process gets picked and how its lever is measured. The answer shows you whether the work starts with a specific workflow in your business or with a general plan.
Our answer: It all starts with a map of every recurring process in your business: who does what, how often, and how long it takes. We rate every process by time, money, risk and feasibility. From that, we pick the process with the biggest lever on time, money or revenue. By day 14 at the latest, we lock in your process in writing.
2. How do you measure that the process is running?
"It's going well" isn't a yardstick. Ask for a definition you can check yourself, and ask who keeps track of where things stand.
Our answer: We measure every process against four autonomy levels. At Level 1, Assisted, the AI employee does its task when someone starts it, and your team checks every run. At Level 2, Autopilot, it starts on its own on a schedule or trigger, runs through to the final report and hands anything uncertain to a person. Level 3 is called Audited, Level 4 Autonomous. Once a month, we look at the level of every process together.
What all four levels mean is explained in The four autonomy levels: how a process in your business runs on its own.
Questions about control
3. Who checks the runs?
This is where a demo and day-to-day operation part ways. An AI system that runs flawlessly in a demo can still make mistakes in daily work. Mistakes happen, especially at the start. The question is who notices them.
Our answer: That depends on the level. At Level 1, your team checks every run. At Level 2, you get a report after every run, and anything uncertain goes to a person. At Level 3, an AI manager reviews the work and bundles the reports. In Kevin's own business, this reviewing role is still on trial. That's why we don't sell it as a finished solution.
4. What happens when the AI is unsure?
This question matters more than any question about the model. A system that guesses when it's unsure needs someone to look over its shoulder for good.
Our answer: With us, every AI employee has an escalation path. If it's unsure, it sends the case to a person on your team. In a feedback session, the AI employee is trained on the new variant, and it's saved as a rule. At the tax advisory firm in our example, anything the filing employee can't match with certainty goes to the office team by email.
5. Who on my team is responsible?
An AI employee without a contact person belongs to nobody. Then every question lands back with you.
Our answer: You name one person on your team who looks after the AI employees with you. That person joins you in the weekly 1:1. The job description of every AI employee names a contact person on your team, along with its role, tasks, permissions and limits.
6. What do I get as a report?
Ask for an example and have them show you a sample. A real report tells you faster than any description what you'll be reading every day later on.
Our answer: After every run, a completion email: what's done, what's still open, where there's an exception. Without a report, you end up following up yourself, and you're back to managing every step. Once several processes are running, the next goal is a morning report in which an AI manager bundles the reports and flags exceptions.
7. What stays with people?
Ask at which points a person deliberately stays involved, and why at exactly those points.
Our answer: For every process, we ask: thinking or doing? Where someone has to think, the work stays with a person. On top of that, there's a house rule: with us, a person stays involved when money, people or legal consequences are at stake. Every step that can't be undone has its own approval in front of it, or the step stays with a person. Typical hard limits in the process profile, our word for a process description, are: delete nothing, send nothing outside the business, make no payments.
Questions about cost, risk and proof
8. What does running it cost once the project is over?
Ask about both: the setup and the ongoing operation afterwards. And ask what belongs to you when the work together ends.
Our answer: We don't give a figure here. We go over how the program works and what it costs on the call, once it's clear which process would come first for you. What ongoing operation costs depends on where the AI employees run in your business and which models they use. We decide that together in the first sessions. Templates and code stay with you, and there's no minimum term.
9. What if it doesn't work out?
An honest question deserves an answer with conditions. Ask what exactly is promised and what you have to do in return.
Our answer: Our guarantee covers your process, and it reads exactly as on our home page: if your process isn't running on Autopilot (Autonomy Level 2) after 90 days, Kevin will keep working with you at no extra cost, as long as you come to the 1:1 every week, complete the agreed actions and name one person on your team.
Autopilot means: the process starts on its own, on a fixed schedule or triggered by an event, and runs through to the final report without a nudge. Anything beyond this guarantee, we call a goal.
10. What do you have running in your own business?
This is the simplest question and often the most telling one. Anyone selling AI employees should be running some themselves.
Our answer: Kevin Welter runs his business with 14 AI employees. What they do, and what is still on trial there, is described below in the section "What this looks like for us".
What to do with the answers
You don't need to be able to judge every answer on its technical merits. Look out for three things.
Is the answer concrete? A process with a name, a level with a definition, a person with a role. General sentences about "potential" don't answer any of the ten questions.
Is there a limit? For question 7, something should be named that stays with a person. If that answer is missing, ask again.
Is there proof? A running process, a client example, the provider's own business. Proof can be small. It should be real and have a date.
If you want to see what proof like that can look like, read AI for SMEs: two processes that really run on their own in 2026, and what it took. And if a provider wants to sell you a ready-made AI system instead of building a process with you, AI agents for business: vendor, platform or your own AI employee? helps you sort out the options.
How Kevin himself looks at processes in small and mid-sized businesses, he describes on his own site: Process Optimization for SMEs.
What this looks like for us
The ten answers above aren't theory. They come from two places.
Kevin's own business. Kevin Welter runs his business with 14 AI employees: content, SEO, video editing, bookkeeping prep, leads and email. Each one has a personnel file, a fixed role and clear limits. An AI manager reviews their work; that role is still on trial. Kevin tries new levels in his own business first, before they come to yours.
A measurable example: the website kevinwelter.com is looked after by an AI SEO employee. In 28 days (Sep 2 to Sep 29, 2026), it got 1,942 clicks from Google Search; in the 28 days before that, it was 88.
A tax advisory firm. Two processes run there at Autonomy Level 2, Autopilot, as of September 2026: filing the digital assessment notices and the notices that arrive on paper. It took 2 weeks there from the first round of watching to sign-off. That's what happened in this one case, not a time promise for your business. Anything the AI employee can't match with certainty goes to the office team and gets trained in afterwards.
What we don't have: a process at Level 4, Autonomous. None of our processes has passed the week-away test yet, not even in Kevin's own business. If a provider tells you their AI runs without any oversight at all, ask since when and how they measure it.
Runs without me
Every two weeks, one process that runs without the owner: what came in, what the AI did, where a person decided. Two minutes to read.
Worksheet
AI Rulebook on One Page
What your AI employee may do, what it must never do, who keeps an eye on it and what is checked before release: eight fields to fill in and an acceptance checklist. Comes as a PDF by email once you confirm.
Strategy session
In a 20-minute strategy session, you tell us where things get stuck. We'll tell you honestly whether an AI employee is a good fit for you.
Book a strategy sessionFrequently asked questions
Do I need an AI strategy before I talk to a consultant?
Not necessarily. More helpful than a strategy is an honest list: which work comes up in your business every day, who does it and how long it takes. That list turns into the map, and the map turns into the first process.
How much time does working with you take?
A 60-minute 1:1 with Kevin every week. Before the next 1:1, you and the person on your team complete the agreed tasks. How much that is depends on the process.
Where does my data stay?
Your documents stay in your own filing system. We don't run servers for you; we help you set up the AI employees in your business. In the first sessions, we decide together where they run and how far that can be set up in line with data protection rules.
What if a provider has no answer to one of the questions?
That's not a reason to rule them out yet. Some questions can only be answered once the first process is settled. Then ask when the answer will be ready and what it will be based on. An answer with a date is helpful, for example: "We'll settle that with you in the first two weeks." That's how we do it too: we lock in your process in writing by day 14 at the latest.
Who is your offer not meant for?
Solo business owners without a team, businesses that aren't up and running yet, large corporations, and anyone looking for a general AI course. Owner Autonomy is meant for businesses with a team of roughly 3 to 15 employees.
Your next step
Our answers above are short. What they look like for your business only shows on your own process. Bring the ten questions to the call, and we'll answer them for the workflow that would come first in your business. There's more about the program on our home page. We go over how it works and what it costs on the call.