Why isn’t AI saving your business any time?
For most businesses the answer is that AI is speeding up individual tasks while the way the business runs stays exactly the same. Everyone's using AI now. The interesting question is what they're actually doing with it, and for most businesses the brutal answer is: using it like a faster Google.
TLDR:
Businesses create the greatest advantage from AI when they use it to redesign workflows. In New Zealand, only 15% of SMEs are using AI extensively and expecting it to reshape how their business operates, while 44% are still using it task by task.
The advantage comes from the question a business asked at the start: how can AI help us do what we already do, or what would we do differently if this wasn't a constraint?
Start with a task that is frequent, done the same way each time, and low risk if it goes wrong.
Personal productivity skills (writing style, morning briefings, meeting prep) are a good first step. Real advantage comes from joining several small automations into a rebuilt workflow.
How are businesses using AI today?
Everyone has access. Very few have depth.
Surface-level AI use looks like this: writing emails faster, summarising documents, generating a first draft of a proposal. Faster ideation, research and opinion. All of it helps.
This would put you miles ahead a year ago. None of it puts you ahead today though, because every business in your market has the same tools and is doing the same things with them.
The businesses pulling ahead have gone past individual tasks and built AI into how the business runs. Workflows that used to take a person three hours now run in minutes. Decisions that required a manager get made at the point of contact. Client reporting that was a Friday afternoon job goes out automatically on Thursday morning.
That's a structural shift in what a business can do with the same headcount.
Two businesses, same tool, completely different outcomes.
We've seen this play out firsthand in NZ. Two businesses in similar markets, both using AI tools, both paying roughly the same for access.
One is using it to write better emails, reviews, proposals and social content. Their team is grateful for the help and their output looks more polished and considerably faster.
The other rebuilt their client intake process. What used to take a full admin day now takes 45 minutes. Then they put the freed capacity into taking on two more clients a month, then systematised that too. The outcome: centralised information gathering, process workflow and customised output all automated via scheduled tasks or a prompted skill.
Same tool, same price. One of them has faster, smarter-sounding content while the other has two extra clients a month and growing capacity.
The gap between them is not the software. It's the question they asked when they started. Business one asked how AI could help them do what they already do. Business two asked what they'd do differently if this wasn't a constraint anymore.
The gap is already measurable.
MYOB published numbers on this in August. Across just over 1,000 New Zealand business owners and managers, 64% of businesses with 20 or more staff are proactively using AI. Among businesses with one to five staff, it's 30%.
The more revealing number is what happens if you take it away. Among the larger businesses, 92% said they would feel it in workload, productivity or cost if they had to stop using AI tomorrow. Among the smallest, fewer than half would notice.
One group has built AI into how the work happens. The other is still trying it out. That gap will compound.
AI for SMEs is still early. There’s an opportunity there.
The same research shows what businesses are actually doing with it: social media and marketing content (38%), copywriting (28%), writing up documents and reports (26%). It's all content generation, and content generation doesn't change how a business runs.
Xero's May research, covering a nationally representative sample of around 1,100 Kiwi small and medium businesses, breaks the market into groups. It found 44% are what it calls Explorers, using AI for specific tasks and largely teaching themselves as they go. Only 15% are Trailblazers, using AI extensively and convinced it can change how their business operates.
So most NZ SMEs are at the content stage. Roughly one in seven have reached the workflow stage. Almost none are at the business model stage, which is where AI starts to change what you can offer, who you can serve, and what your margins look like.
Xero also found that 64% of owners using AI say it has saved them time, an average saving it describes as equivalent to one working day. That's a real result. It's also the kind of result your competitor can replicate by Friday, which is why it isn't where the advantage lives.
The window to get ahead of this is open right now.
At Point16 we work with businesses across a range of industries. The ones moving fastest on AI are not the most tech-savvy. They're the clearest on where their time actually goes and what they'd change if they could. That clarity is the starting point.
If you don't know where your hours are being lost, no tool will fix it. AI just makes the loss faster.
How to find the process that should be automated.
Most people get stuck when they finish an article like this one, agree with it, and then open their inbox with no idea what to do next.
So let's make it simple. Run every task through these three questions.
Does this task happen every week? A one-off job won't repay the setup effort. You want something frequent enough that fixing it pays for itself quickly.
Do you do this task roughly the same way each time? Tasks that need fresh judgment every time are a poor place to start. Aim for the work you could hand to a new starter with a checklist.
If it went wrong, would you catch it early or would a client see it first? This is the guardrail question, and it matters more than the first two. Start where a mistake is cheap and easy to spot. Anything client-facing, financial, or hard to undo needs a human checkpoint built in.
That third question is the one businesses skip and where they get caught out. The EMA surveyed more than 300 member businesses in July and found 83% are using AI, 13% have any kind of AI policy, and in 18% nobody owns it at all.
So our Point16 test is: frequent + repeatable + low-risk. Resist the urge to start with the biggest problem in your business and pick something boring instead.
Start small on purpose.
There's a fair objection here. We've spent half this article saying task-level AI use won't get you ahead, and now we're recommending you start with small, boring tasks.
Both hold, and the sequence is the reason. The businesses that reached the workflow stage started with something small enough that they'd actually finish it. That's where they learned what AI handles reliably, where it needs checking, and which of their processes was worth rebuilding. Skipping that step is how you end up six weeks into redesigning a core process with nothing running.
Start small, and keep sight of where it leads. The first few skills are practice.
How to set up your first one.
Here's how you actually build one. The tool for the job is a skill: a pre-configured instruction set that tells your AI what to do, and when, for one specific process. You build it once, then refine it as you go, feeding it corrections until it fits the way you work.
We build ours in Claude, though the same approach works across most of the major tools.
One thing before you start. Each of these connects AI to your email or calendar. Use a business account where you control the admin settings, check your provider's data settings so your content isn't used to train models, and keep access read-only until you're confident enough to let something send on your behalf.
Here are a few of our team's favourites:
Your writing style
Once this skill is set up, any time you ask AI to draft an email or message it writes in your voice: your tone, your length, your habits. You never have to explain how you write.
Connects to: your email account (sent folder)
What it does: reads your sent emails, identifies your writing patterns, and builds a permanent style guide. Every future draft matches your voice automatically.
How to set it up: connect your email. Ask your AI tool to read your last 50 sent emails and write a style guide covering your tone, sentence length, how you open and close, and what you never say. Save it as a skill, or in your instructions.
Your daily morning briefing
Every morning you get a short brief covering what's on today, anything you need to follow up from yesterday's email, and anything worth flagging. You open it first thing instead of opening your inbox. Some of our team have configured theirs to pull in news articles and podcasts too.
Connects to: calendar and email
What it does: runs automatically each morning. Pulls today's meetings, summarises key emails requiring action, and flags anything time-sensitive.
How to set it up: set up a scheduled task that runs at a time that suits you each morning. Tell it to check your calendar for today and scan your email for anything requiring action or follow-up.
Weekly team meeting prep
Before your weekly team meeting you get a structured summary of last week and this week. Takes the prep work out of running the meeting.
Connects to: calendar and email
What it does: runs every Monday morning. Summarises last week's key meetings and any commitments made in sent email, then lists this week's meetings and open actions.
How to set it up: set up a scheduled task to run Monday mornings. Tell it to look back over the past week's calendar and sent emails, then look ahead at the week's calendar, and format the output as a structured WIP brief.
What stage two looks like.
Those three are personal productivity. They'll give you back hours in your week, and they'll leave your business fundamentally unchanged. Worth doing as a first step.
Here's what the next stage looks like in practice. That client intake process from earlier, the one that went from a full admin day to 45 minutes, started as three separate skills. One pulled the key details out of an enquiry email into a standard brief. One drafted the onboarding pack from that brief. One checked the pack for gaps before it went out.
Each of them was fairly ordinary on its own. The difference came from noticing that all three were touching the same process, and joining them up.
That's usually how it goes. You build three or four small things, notice they're all pointing at the same bottleneck, and connect them. The workflow emerges.
Start with one, then another, and another.
If AI isn't saving you time right now, that's fine. It can, and the bar to get started is lower than most people think.
The businesses that will look back on 2026 as the year things shifted are the ones who stopped asking how AI could help them do what they already do, and started asking what they'd do differently if it wasn't a constraint.
That question costs nothing to ask. Start there.
If you'd rather not work it out alone
Xero asked Kiwi business owners what support they most wanted. Practical training came first, named by 47%, ahead of access to vetted tools and real-world case studies. Meanwhile 79% of those already using AI are teaching themselves through trial and error.
Point16 runs a practical half-day AI workshop for NZ businesses that want to move beyond prompting and start building AI into how their business actually works. We work with your real tasks and processes, so you leave with a workflow that works from day one.
Sources
MYOB Business Monitor, August 2026, n=1,026 NZ business owners and managers. Coverage via IT Brief NZ
Xero, From curious to confident: how Kiwi business owners are adopting AI, May 2026, n=1,081. Xero Blog
EMA member survey, July 2026, n=300+. EMA
Frequently Asked Questions
How do I know if my business is actually behind on AI, or just using it differently?
If your AI use is limited to writing content faster (emails, proposals, social posts) you are in the largest group of NZ businesses, and that group is not gaining ground on competitors doing the same thing. The businesses pulling ahead have used AI to change a process. A simple test: did AI remove a step from how your business runs this month, or did it just make an existing step faster?
What should I automate first if I don’t know where to start?
Pick a task that happens weekly, is done roughly the same way each time, and is low risk if something goes wrong. Avoid starting with your biggest or most client-facing problem. Small, repeatable, boring tasks are the right entry point because you will actually finish setting them up, and that is where you learn what to trust AI with.
Is it risky to connect AI tools to my email or calendar?
It carries some risk if set up carelessly, which is likely why oversight is lagging adoption nationally (one recent NZ survey found the large majority of businesses using AI, but only a small fraction with any AI policy). Use a business account you control, check your provider’s data and training settings, and keep access read-only until you trust the setup enough to let it act on your behalf.
How long does it actually take to see a result from this?
Personal productivity tools (a writing style guide, a morning briefing, meeting prep) can be set up quickly and save time within the first week. Bigger wins, like rebuilding a full workflow, come later and usually grow out of two or three smaller tools you already built, once you notice they’re touching the same process.
Note: this article was written in August 2026. The way AI is going, some of it will be out of date within weeks, so keep that in mind if you're reading it later in 2026 or beyond.
