Cloud & AI guides
Cloud servers, Copilot and using AI safely at work. 25 plain-English guides for small and medium businesses.
- What "the cloud" actually means for a small businessThe cloud simply means software and storage that run on someone else's computers in a data centre, which you reach over the internet and pay for by subscription. Most small offices already use it every day through email, accounting and file sharing, so the real question is not whether to use it but how to run it well.
- Cloud or onsite server: comparing the real costs and risksFor most offices under about 50 people, moving files and email to cloud services costs about the same as buying a new server over five years but removes most of the hardware risk. An onsite server still makes sense when you run software that needs one, have very large files on a slow connection, or must keep data physically on premises.
- How to move an old office server to the cloud without chaosMoving an old server to the cloud works best as several smaller moves: files to SharePoint or a similar service, user sign-in to a cloud directory, and any remaining application to a hosted server or its vendor's own cloud version. Plan it around what the server actually does today, not around the hardware.
- Microsoft Azure for small business, in plain EnglishAzure is Microsoft's rent-a-data-centre service: you hire virtual servers, storage, databases and networking by the hour or month instead of buying hardware. Microsoft 365 runs on top of Microsoft's cloud too, but Azure is the part where you build and manage things yourself, so most small offices only need a small piece of it.
- Azure Virtual Desktop and hosted desktops: when they make senseA hosted desktop runs Windows and your applications on a server in a data centre, and staff connect to it from any device. It makes sense when people must use a server-based application from many places, when devices need to stay free of company data, or when staff come and go often; for an ordinary office on Microsoft 365, a well-managed laptop is usually simpler and cheaper.
- Running line-of-business software that still needs a serverIf a key business application only runs on a Windows server, you have four realistic choices: keep a small server onsite, rent a hosted server in a data centre, move to the vendor's own cloud edition, or replace the software. The right one depends mostly on what the vendor supports and how many people use it away from the office.
- Why your cloud bills keep creeping up, and how to control themCloud costs creep because each extra user, add-on, storage tier and forgotten trial adds a small amount that nobody reviews. A quarterly routine of listing every subscription, matching licences to real people and checking usage usually finds savings within an hour or two.
- SaaS sprawl and shadow IT: finding the apps nobody approvedShadow IT is any app or online service staff use for work without the business knowing, and SaaS sprawl is what happens when dozens of them pile up. The fix is not banning everything; it is finding what is in use, keeping the useful tools under proper company accounts and giving people an easy way to ask for new ones.
- Choosing cloud software: the security questions to ask the vendorBefore you put client or staff data into a new cloud product, ask how the vendor protects sign-ins, where data is stored, how it is backed up, who can see it, what happens in a breach and how you get your data out if you leave. Good vendors answer these quickly and clearly; vague answers tell you something too.
- Data sovereignty: where your business data is stored and why it mattersData sovereignty means data is subject to the laws of the country where it is stored, so data held in Australia or the United States can be reached under those countries' rules as well as ours. For most NZ small businesses, offshore storage with a reputable provider is acceptable, but you should know where your data lives and check any contracts that say otherwise.
- Single sign-on across your cloud apps: what it is and why it helpsSingle sign-on lets staff log in to many cloud apps with their one work account, usually Microsoft or Google, instead of a separate password for each. It improves security because there is one strong, protected sign-in to manage, and it makes leavers far easier to handle because switching off one account closes every app at once.
- SharePoint, Google Drive or Dropbox: comparing cloud file storageAll three store and sync business files well; the right choice usually follows the email and office apps you already use. SharePoint and OneDrive fit Microsoft 365 offices, Google Drive fits Google Workspace offices, and Dropbox suits teams that want a simple, platform-neutral file service or work heavily with outside collaborators.
- AI at work: practical uses for a small businessAI tools are most useful in a small business for first drafts, summaries, reformatting information and answering questions about your own documents, saving minutes on many everyday tasks rather than replacing whole jobs. They work best when a person who knows the subject checks the output before it goes anywhere important.
- Microsoft Copilot or ChatGPT for business: how to chooseMicrosoft 365 Copilot is built into Word, Excel, Outlook and Teams and can draw on your company's own emails and files, while ChatGPT's business plans are a strong general assistant you bring information to. Offices deep in Microsoft 365 often lean towards Copilot; others value ChatGPT's flexibility, and some use both for different jobs.
- AI risks at work: confidential data in prompts and confident wrong answersThe two AI risks that most often hurt small businesses are staff putting confidential information into the wrong tool, and staff trusting answers that sound right but are false. Both are reduced by giving people a business-grade tool, clear rules about what can be entered, and the habit of checking anything important before it is used.
- How to write an AI use policy your staff will actually followA good AI use policy for a small business fits on one or two pages and answers four questions: which AI tools are approved, what information must never go into them, which outputs need a person to check them, and who to ask when unsure. Write it in plain language with examples from your own work, and review it every six months.
- Preparing your data and permissions before you switch on AIAI assistants connected to company files will find and repeat whatever they can reach, including out-of-date documents and folders shared too widely. Before switching one on, archive old material, fix who can see what, label the sensitive information and make sure each important folder has an owner.
- Using AI for email, meetings and documents day to dayUsed well, AI assistants can trim time from three daily chores: working through email, capturing meeting notes and actions, and drafting or reviewing documents. The gains come from clear instructions, giving the tool the right context and a quick human check before anything leaves the building.
- Automating routine office tasks with Power Automate and similar toolsWorkflow tools such as Microsoft Power Automate, Zapier and Make connect your cloud apps so routine steps happen automatically: saving attachments, sending reminders, routing approvals and copying data between systems. Start with one small, frequent, rule-based task, document it, and make sure someone owns each automation.
- AI-generated scams: deepfake voices, fake video calls and convincing fake invoicesAI lets scammers clone a manager's voice from a short clip, appear on a video call as someone else and write invoice emails with no spelling mistakes in your supplier's style. The defence is not spotting the fake, which is getting harder, but having a firm rule that payments and sensitive requests are verified through a separate, known channel every time.
- How to evaluate the AI tools vendors pitch to your businessJudge an AI product by a specific problem it solves in your business, tested on your own work, not by a demo. Ask what data it uses and where that data goes, what it costs once the trial ends, how you will know it is accurate and what happens to your information if you stop using it.
- AI and the Privacy Act 2020: what NZ businesses need to knowNew Zealand's Privacy Act 2020 applies whenever a business uses AI with personal information, just as it does with any other software. In practice that means only putting personal information into AI tools for purposes people would expect, keeping it secure, checking AI output about people for accuracy and being able to explain what you did with it.
- How to train staff to use AI tools wellGood AI training for a small business is short, practical and based on each person's real tasks: how to give clear instructions, what information to keep out, how to check results and where AI is not the right tool. Follow it up with shared examples and a quick refresher every few months, because the tools keep changing.
- When a cloud service goes down, or a vendor disappearsEvery cloud service has outages from time to time, and smaller vendors are occasionally sold or shut down. Prepare by knowing which services you cannot work without, how staff will carry on for a few hours, where to check status, and how you would get your data out if a vendor closed.
- AI agents explained: when software starts acting on your behalfAn AI agent is an AI tool that does not just answer questions but carries out steps for you, such as reading an inbox, updating records, booking meetings or placing orders, often across several systems. Agents can save real time on routine work, but because they act rather than advise, they need tight limits, logs and a person approving anything that matters.