AI or more staff: how do you make that choice?

The question that keeps coming up

Labour shortages have been a persistent problem for many Dutch business owners for years. Vacancies remain unfilled, good people are hard to find, and wage costs keep rising. Meanwhile, everyone seems to be saying AI is the answer. But is that actually true?

The honest answer: AI is not a replacement for people. It is a tool that performs specific tasks better, faster and more consistently than a person doing them manually. The real question is not “AI or staff”, but: which tasks are suitable for automation, and which are not?

What AI is good at, and what it is not

At this point in time, AI is strong at tasks that:

• Are repetitive and follow a fixed pattern
• Involve processing or searching large amounts of data
• Require fast responses, including outside office hours
• Are based on existing information (text, data, rules)

Think of lead follow-up via WhatsApp, answering frequently asked questions, processing forms, or sending appointment reminders. These are tasks where a person is mainly spending time on something that requires no human judgement.

AI is less suitable for tasks that:

• Require complex emotional attunement
• Involve unique situations that differ significantly from previously seen patterns
• Centre on accountability and relationship management
• Are legally or ethically sensitive

Replacing employees with AI in those areas yields little, and research confirms this repeatedly. But that does not mean AI adds nothing. It means you need to be more precise about where you deploy it.

Where are time and revenue leaking right now?

Before deciding whether AI or a new employee is the right choice, it is worth taking a step back. Look at your own business:

• Which tasks are performed manually every single day?
• Where do customers get stuck or drop off?
• Which questions do you receive over and over?
• Where are you responding too slowly to leads or customers?

That analysis is worth more than any tool demo. At NRL Automations, we always start there. Not with the technology, but with what your business actually needs. Only once it is clear where time and revenue are leaking does it make sense to look at which solution fits.

Concrete examples of AI relieving the pressure

Automatic lead follow-up

A lead that comes in through a form or advertisement waits too long for a response on average. An AI agent can respond within seconds via WhatsApp or email, ask qualifying questions and schedule an appointment. No employee needed for that first step.

Customer service outside office hours

An AI agent built on your own knowledge and documents answers frequently asked questions, directs customers to the right information and escalates to a staff member when things get too complex. That is exactly how you combine AI and human contact well: the agent handles the groundwork, the employee does what actually matters.

Administrative processing

Processing forms, forwarding data to the right system, sending appointment reminders, following up on invoices: these are tasks that can be automated perfectly well using tools like n8n and integrations with your CRM, Stripe or other software.

Reports and overviews

Instead of manually gathering data from various tools every week, a tracking dashboard gives you real-time insight into what is happening. That not only saves time, it also provides better information to base decisions on.

When is hiring more staff the right answer?

AI does not solve people problems. If you are growing in client relationships, complex service delivery or creative work, you need people. The question then is: are those people currently doing work that truly requires their attention, or are they also handling tasks a system could take over?

When you automate routine tasks, existing employees get room for the work that really matters. That is a stronger foundation than hiring someone for tasks you could actually automate.

The article on automating business processes: where do you start walks you through that analysis step by step.

What is happening technically right now?

The current generation of AI agents, built on models like Claude from Anthropic, has become considerably better at understanding context and making small decisions within a defined process. Coupled with your own data via RAG (Retrieval Augmented Generation), tools like n8n for workflow automation and integrations with platforms such as GoHighLevel or Supabase, you can build systems that genuinely work in practice.

It is no longer a question of whether AI is good enough. It is a question of whether you are giving the right task to the right system, and whether that system is set up properly and securely.

The balance in practice

The businesses that benefit most from AI are not the ones deploying it at the largest scale, but the ones deploying it most intelligently. That means:

• Starting with a concrete problem, not with a technology
• Beginning with one process that you automate and measure
• Making sure AI and human work complement rather than compete with each other
• Setting everything up securely from day one, with proper access control and data management

Deploying AI during a labour shortage is smart if you know what you are using it for. It becomes a problem if you expect too much from it or hand over tasks that require human judgement.

Ready to see where it fits for you?

If you want to know which processes in your business are suitable for automation, and what that concretely delivers, a conversation is the fastest way to get clarity. Plan a conversation and we will look together at where the quick wins are.