Hiring an automation specialist: what to ask
When you hire an automation specialist, whether a freelancer, an agency or an employee, ask them how they work before you ask what they cost. The useful questions are about how they find out what you need, what they build, how they test it, what happens when it breaks, and who owns the result once they leave. This guide gives you the questions and what to listen for in the answers.
Good specialists ask you questions first, talk about exceptions and failure, and are clear about ownership. Be wary of anyone who promises a result before they understand your process, or who cannot explain what you would be left with at the end.
What should you prepare before talking to anyone?
You will get better answers if you can describe what you want. Write a short brief: the process you want help with, what starts it, which tools are involved, what goes wrong today and who would use the result. A rough page is enough. If you have not done this, documenting the process before you automate it shows how, and the automation readiness checklist will show whether you are ready to start.
Decide too which kind of help you need. A one-off build, ongoing support and advice on what to automate are different jobs, and some people are better at one than the others.
Questions about how they start
- "How would you find out how my process works today?" A good answer involves talking to the people who do the work, looking at real examples and asking about odd cases. A weak answer jumps straight to tools.
- "What would you need from me?" Access, examples, a named contact and decisions. This tells you how much of your time it will take.
- "What would you not automate in this process?" Someone who can name parts to leave with a person understands the risks. Someone who says "everything" may not.
Questions about how they build
- "How do you choose the tools?" Look for reasoning based on your needs and team, not a single favourite for every job. Our guide to Zapier, Make or n8n lists the trade-offs they should be able to talk about.
- "How do you handle exceptions and bad data?" They should mention checks, routing to a person and logging. See the exception problem and designing a review queue.
- "Will a person approve things at first?" Starting with approval and relaxing later is a sensible pattern.
- "How do you keep things secure?" Ask where your data goes, who has access and how credentials are stored. You do not need to be an expert; you need to hear a clear answer.
Questions about testing and launch
- "How will we know it works before it goes live?" Listen for testing with real examples and running beside the manual process, as in the parallel run.
- "What does success look like, and how will we measure it?" They should be willing to agree a simple measure with you; see measuring automation results.
- "What can go wrong at launch?" An honest person will have a list.
Questions about what happens after
This is where many arrangements fall down. Automations depend on other systems that change, and they do break; see why automations break. Ask:
- "Who gets alerted when something fails, and what happens then?" You want alerts and a named person, not hope.
- "What is your support arrangement, and what does it cover?" Get it in writing, including how changes are requested.
- "Who owns the accounts and the work?" The automations should live in accounts you control, or you should know exactly how to take them over.
- "What documentation will I get?" You should receive a description of what was built, how it works and what to do when it fails.
- "What happens if you are unavailable?" A single person with sole access is a risk worth knowing about.
How do you check their experience honestly?
Ask them to describe a past project in some detail: the problem, what they built, what went wrong and what they changed. Ask for references you can contact, or examples you can look at, where confidentiality allows. Real experience usually comes with specific stories about mistakes. Be cautious of anything that sounds too smooth, and remember that a polished portfolio does not tell you how they handle problems.
You can also test them with a small, paid first step, such as mapping one process or building one limited workflow, before committing to something larger. See what a starter automation gets you for what a small first project can look like.
What are the warning signs?
- Promises of specific results before they have seen your process.
- No questions about exceptions, data quality or who will own it.
- Reluctance to put scope, support and ownership in writing.
- Everything is built in accounts only they can access.
- Pressure to decide quickly or sign a long commitment before a small trial.
- A solution chosen before the problem is understood, such as "you need an AI agent" in the first call.
On that last point, our article on AI automation or a new assistant helps you decide which kind of help fits.
Frequently asked questions
Should I hire a freelancer, an agency or an employee?
It depends on how much ongoing work you have, how much support you need and how much management time you can give. The questions above apply to all three.
Do I need to understand the technology?
No, but you should understand your own process and be able to judge whether the answers are clear. If you cannot follow an explanation, ask for it in plainer terms.
Is it a problem if they use no-code tools?
Not in itself. The question is whether the solution is reliable, documented and maintainable.
Should I ask for a fixed scope?
For a defined first project, yes. For ongoing work, agree how changes are requested and handled.
Want to see how we would answer these questions? Book a free audit and ask us directly.
Want this built for your business?
We will look at how your work flows today and tell you which steps are worth automating, and which are not. Free, and you keep the notes.
![]()
You will talk to Fayez or Kareem, the two founders.
