The New Role That Confuses Everyone
Something interesting is happening in small business hiring. Job boards and LinkedIn are filling up with titles like "AI Automation Specialist," "Workflow Automation Consultant," "Prompt Engineer," and "AI Integration Lead."
Some of these people are genuinely valuable. Some are skilled operators who've learned a few n8n workflows and rebranded. And some are charging £500 a day to do things you could learn in an afternoon.
The question isn't whether to take AI and automation seriously — you should. The question is whether bringing in someone else to do it is the right move for your business right now. For some businesses it clearly is. For many, it isn't yet.
Here's how to figure out which camp you're in.
What This Role Actually Does
Before you can decide whether to hire, you need to understand what you'd actually be getting.
A competent AI/automation person does some combination of:
Workflow automation. Building the pipes that connect your tools together — Zapier, Make, n8n, or custom scripts. When X happens in tool A, trigger Y in tool B. They set this up, document it, and maintain it.
AI integration. Embedding AI into your existing processes — adding an AI drafting step to your client comms, setting up Claude or GPT as part of a workflow, building custom AI assistants with business-specific context.
Process mapping. Figuring out which of your processes are actually worth automating, in what order, and what the ROI looks like. Good operators do this before they build anything. Bad ones just build things and invoice you.
Maintenance and iteration. Workflows break. Tools update their APIs. Prompts that worked six months ago stop working. An ongoing relationship means someone's managing this. A one-off consultant means you're on your own.
The range in quality and scope is enormous. A freelance n8n specialist and an AI strategy consultant are both calling themselves "AI/automation people" right now — but they're doing very different work and worth very different amounts.
The Case for Hiring
There are situations where bringing someone in genuinely makes sense.
You've done the audit and you know what needs building. You've identified three high-value workflows, you know what the inputs and outputs are, and you just need someone with the technical skills to build them faster than you could. A contractor who can come in, build cleanly, document it well, and hand it over is worth paying for.
Your time is expensive. If your hourly rate is high and learning automation tools would take you 30+ hours to reach competence, the maths can work out in favour of hiring. Time-to-value matters. If you need this running in two weeks and you haven't started learning, hire.
The complexity is genuinely high. Multi-system integrations, custom APIs, complex conditional logic, data pipelines — these take real technical skill. Not every automation job is drag-and-drop Zapier. If what you need is legitimately technical, a skilled person will do it better and faster than you learning on the job.
You've tried DIY and got stuck. You built something, it works 80% of the time, but the 20% that breaks is eating your time. A professional can often diagnose and fix what a beginner set up poorly in less time than it would take you to figure it out yourself.
The Case for Doing It Yourself
For most small businesses at early AI adoption stages, DIY is usually the right call.
You know your business better than any consultant will. The context needed to build a good workflow — your processes, your clients, your edge cases, your quirks — lives in your head. A contractor has to extract that from you in discovery sessions. You already have it. That's a real advantage.
The tools are genuinely learnable. Zapier and Make are drag-and-drop. n8n has a learning curve but it's not steep. Claude and ChatGPT prompting for business use can be picked up in a weekend. If you can commit 10–15 hours to learning, you can build your first workflows yourself.
You'll maintain it better. Workflows need updating. A system you built yourself is one you understand. A system a contractor built, handed over with a brief video, and left — is a system you'll be afraid to touch when something breaks. Ownership matters for longevity.
Most businesses aren't ready for a specialist. If you haven't identified specific workflows, don't have a clear sense of your highest-value automation opportunities, and don't know what success looks like — you're not ready to hire. You'd be paying someone to figure out the strategy while you pay them to do it. Do the audit first (there's a guide on this site for that). Then you'll know if you need to hire.
The Test That Settles It
Answer these four questions honestly:
1. Can you describe exactly what you need built? Not "I want to automate my business." Specifically: what is the task, where do the inputs come from, what should the output look like, where should it go. If you can't answer this, you're not ready to hire — a consultant will spend your money figuring it out.
2. Would it take you more than 40 hours to learn to build this? If yes, and your time is genuinely worth more than £15–25/hour, hiring may make financial sense. If no, learn it yourself and own it.
3. Do you have ongoing budget for maintenance and updates? A one-off automation build that no one maintains will be broken within six months. If you can't commit to either maintaining it yourself or paying for ongoing support, factor that into your decision.
4. Is there a contractor you can vet properly? Have you seen their actual work? Previous clients you can speak to? A clear brief they've responded to specifically? If you're hiring from a cold LinkedIn DM with no references, you're rolling the dice.
If you answered yes to all four: hire. If you answered no to any of them: not yet.
If You Do Hire: What to Look For
The title means nothing. Look at the work.
Ask to see three workflows they've built for businesses similar to yours. Ask what broke and how they fixed it. Ask how they document their work. Ask what tools they use and why. If the answers are vague, that's your answer.
Avoid anyone who:
- Pitches you on automation before asking questions about your business
- Uses "AI-powered" as a selling point without specifics
- Quotes you a high day rate for "strategy" without a clear deliverable
- Can't show you working examples
Look for someone who:
- Does a proper discovery session before proposing anything
- Builds with documentation as part of the process, not an afterthought
- Is honest about what needs to be maintained and what the ongoing cost is
- Hands over in a way that leaves you less dependent on them, not more
The best automation contractors make themselves unnecessary. If someone's incentive is to keep you dependent on them, that's a problem.
If You DIY: Where to Start
Start with the AI audit. One day of tracking your time, two weeks of end-of-day logs, then an hour identifying your highest-value repeatable task. That's your first automation target.
For building: start with the tool that's lowest friction given your existing stack. If everything's in Google Workspace, Make or Zapier is the fastest path. If you're technical and want more control, n8n is worth the extra setup time.
Build one thing. Use it for 30 days. Measure what changed. Then build the next one.
The learning curve is real but it's not long. Most people who commit a serious 10 hours to learning Zapier or Make can build something genuinely useful by the end of it. What takes months is building the habit of reaching for these tools when the right opportunities come up.
That habit is worth more in the long run than any one contractor you could hire.
The Honest Answer
Most small businesses reading this: not yet.
Do the audit. Learn the basics. Build your first two or three workflows yourself. Once you have working systems and a clear picture of what needs to go further than your current skills — that's when hiring makes sense.
The exception is if you're genuinely time-poor, technically averse, and can afford to pay for someone to shortcut the learning phase. In that case, hire carefully, vet hard, and make sure you end up with documentation you understand.
But for everyone else: the fact that you're reading this suggests you're capable of doing more of this than you think. Start there.