AI is genuinely useful for business right now. You don't need to be technical to benefit from it, and you don't need to overhaul how you work.
What you do need is a clear picture of what's actually on offer - because "AI" now covers several quite different things, and they're showing up in your business in different ways.
One more thing before we start: if you're feeling anxious about falling behind, you can relax a little. Recent global research has found that most large corporate AI implementations have delivered no discernible productivity benefit. That's not because AI doesn't work - it's because large organisations tend to roll it out without clear use cases or practical habits to back it up. The businesses getting real value are the ones starting small, staying practical, and building from there. That's exactly the approach this guide recommends.
The three ways AI shows up in your business
1. AI built into apps you already use
This is the most common form - and many people are using it without thinking of it as AI at all.
Canva, Microsoft Word, Excel, Outlook, and many other everyday tools now have AI features built in. In Canva it can generate images or suggest layouts. In Word it can draft or rewrite text. In Outlook it can summarise long email threads.
Where to start: Look for AI or "magic wand" features inside tools you already have open. They're often useful for small tasks and require no setup.
2. AI assistants - ChatGPT, Claude, Copilot, and others
These are tools you interact with directly, in plain English, to get work done.
ChatGPT (from OpenAI) is the one most people started with, and it remains excellent for a wide range of tasks. Claude (from Anthropic) has emerged as a strong alternative - and is the tool the GTB team now uses as our primary AI assistant (more on that below). Microsoft Copilot is Microsoft's AI assistant, embedded directly inside Outlook, Word, Excel, and Teams — so rather than switching to a separate app, it works right where you already are.
What AI assistants are good for:
Drafting emails, proposals, reports, and job ads
Summarising long documents or email threads quickly
Researching topics and getting plain-English explanations
Thinking through a business problem
Creating first drafts of policies, procedures, or marketing copy
One firm rule: Use a paid licence for business, not the free version. Free tiers may use your data for training, have limits that cut you off at the wrong moment, and aren't designed for business information. Paid plans for ChatGPT and Claude start at around NZ$30–35 per person per month. Copilot is available as an add-on to Microsoft 365.
3. AI that takes action - 'agentic' AI
Up until recently, AI tools answered your questions — and then you did the work. Agentic AI goes further: you describe what you want to achieve, and it plans and carries out the steps to get there, working across apps and systems without you managing each step.
This is still early-stage for most small businesses, but it's developing quickly. It's also where Copilot and other tools are actively heading.
A closer look at Microsoft Copilot
For businesses that work heavily in Microsoft 365, Copilot is worth a serious look. Because it works inside Outlook, Word, Excel, and Teams - rather than as a separate tool - it fits naturally into existing workflows.
Practical examples:
Summarise a long email thread in seconds
Draft a document or email from a few bullet points
Ask Excel to explain what your data means in plain English
Get a meeting recap with action items after a Teams call
A number of our clients have taken on Copilot licences and are finding real value, particularly around email and document work. We can help you work out whether it's the right fit and get it set up securely for your team.
See our Copilot guide for Kiwi SMEs →
A word on keeping your information safe
As AI use becomes more common, it's worth being clear with your team about what should and shouldn't go into these tools.
As a general rule: don't enter client data, confidential business information, or financial details into free or consumer AI tools. Paid business licences come with appropriate terms around data use - free versions do not. If your team is starting to use AI independently, a simple one-page guideline on what's appropriate goes a long way.
Where the skills are going - and what that means for your team
When AI first arrived in most workplaces, the focus was on prompting - learning how to phrase questions well to get better answers. That's still a useful skill.
But increasingly, the people getting the most value from AI are developing a different capability: managing AI to achieve goals rather than just answer questions. Rather than asking "can you help me with this?", they're thinking "what outcome do I need, and how do I get AI to work towards it?" - setting up projects, building workflows, and using AI tools in ways that reduce manual work across whole processes, not just individual tasks.
For most SME teams right now, confident prompting is still the right starting point. But it's worth knowing where this is heading, so that as your comfort grows, the next step isn't a mystery. The businesses that will get the most from AI over the next few years are the ones building these habits now, at small scale, before the tools get even more capable.
Where to start - practical first steps
The most common mistake is waiting until you have a clear "AI strategy." You don't need one. You need a small first project.
Find a repetitive manual process and address it first. Think about tasks in your business that happen regularly, take more time than they should, and follow a predictable pattern. Drafting a standard type of email. Summarising meeting notes. Producing a weekly report from the same data. These are ideal starting points - the AI doesn't need to be perfect, it just needs to save you meaningful time on something you do often. Start there, get comfortable, then look for the next one.
Specific things worth trying this week:
Pick one email you write regularly - a quote follow-up, an onboarding message, a staff update - and ask ChatGPT or Claude to draft it from a few bullet points. Refine the result. Save the prompt that worked best.
Open Word or Outlook and look for the AI features. Ask it to summarise a recent document or email thread.
If you're on Microsoft 365, check whether Copilot is already included in your licence - you may have access you're not using.
Identify one process in your business that involves pulling together information from multiple places and writing it up. That's a strong candidate for an AI workflow.
One thing to avoid: using AI on your most complex, high-stakes work before you've built confidence on simpler tasks. Start where the risk is low and the repetition is high.
What GTB is doing
The GTB team has moved to Claude as our primary AI tool. We're using it across the team - taking the step from strong prompting to company-wide collaboration and standards, using Projects to keep work organised, and learning how to use Skills to tailor how it works for specific tasks, Artifacts and integrations with tools like Excel. We started where most people start (asking it questions), and we're now learning to use it in ways that genuinely change how we work - including building workflows that handle tasks we used to do manually. The aim? Becoming more efficient and focusing our valuable team on value-adding tasks.
We're also helping clients who are invested in Microsoft 365 safely move to, and get real value from Copilot.
What we've found: the biggest barrier isn't the technology. It's getting comfortable enough to start, and then building the habit. We can help with both.
Want to talk it through?
If you're not sure where to start, which tool suits your business, or how to introduce AI to your team sensibly - get in touch. It's exactly the kind of conversation we enjoy.
GTB IT Solutions supports small and medium businesses based in Wellington, Porirua, Hutt Valley, Kāpiti, and into the Horowhenua. Reliable, practical IT - with no confusing jargon.