A plain guide
Putting AI to work in a small business.Five steps, in order.
Most guides tell you which tools exist. This one is about the order to do things in, and where people get stuck. Each step shows how to do it yourself, and what we do if you would rather have help.
Do it yourself, or do it with us.
You can stop after any step. Nothing here needs custom software to begin.
Step 1
Pick one job, not “AI”
AI is not a project. A job is. The businesses that get value start with one piece of repeated work and ignore everything else until that one works.
On your own
Write down the tasks you repeat every week. Pick one that is frequent, dull and low-risk if it goes slightly wrong: sorting enquiries, drafting replies, chasing invoices, summarising documents.
With us
This is what the free scope call is for. You leave with a high-level analysis of where AI fits in your business, where it does not, and a floor price.
Step 2
Choose a tool you can buy today
For most first jobs there is already a tool. Custom software comes later, if it comes at all.
On your own
There are three kinds worth knowing. General assistants, such as ChatGPT, Claude, Copilot or Gemini, for writing and summarising. Connectors, such as Zapier or Make, for moving information between the apps you use. And the AI features already inside software you pay for, which are worth checking first.
With us
In a guided setup we choose the tool with you, from ones you can buy today, and set it up properly. We name tools as examples only and are not affiliated with any of them.
Step 3
Check where your data goes
This is the step most guides skip, and the one that causes real trouble.
On your own
Before you paste a customer's details into any tool, find out where it stores what you give it, whether it uses your data to train its models, and who at the provider can see it. Under privacy laws such as POPIA and GDPR you stay responsible for that information. This is not legal advice.
With us
A check on where your customer data goes is part of every setup, and it happens before any tool touches that data.
Step 4
Keep a person in the loop
The risk with AI is not that it fails. It is that it gives a confident wrong answer and nobody notices.
On your own
Start with drafts, not sends. Have a person read everything the tool produces for the first few weeks. Keep a short log of what you use AI for, so you can see what it is doing in your business.
With us
We build the stop into the workflow. Anything about money, refunds or a commitment is held for a person to approve.
Step 5
Measure, then decide
If you cannot say whether it helped, you cannot decide whether to keep paying for it.
On your own
Before you start, note how long the job takes and how often it goes wrong. Look again after a month. Keep it, change it or drop it.
With us
Every setup includes a before-and-after measure. If an off-the-shelf tool cannot do the job, the next step is a workflow built around how you work.
Stuck on a step?
Tell us which one. The scope call is free, takes 30 minutes, and you leave with a plan whether or not you work with us.
Book a free scope call