It’s the Wi-Fi. It’s always the Wi-Fi.

Bad Wi-Fi isn’t just annoying — it costs you money. Slow EFTPOS terminals mean longer queues. Dropped connections mean staff waste time retrying. And if you’re running any kind of cloud-based system (which most businesses are now), unreliable Wi-Fi means unreliable business.

The good news: you don’t need to spend thousands to fix it. You just need to stop treating Wi-Fi like it’s magic.

Why Small Business Wi-Fi Is Usually Bad

Most small business Wi-Fi setups follow the same pattern:

1. ISP provides a modem/router combo
2. It gets plugged in wherever the phone/fibre line enters the building
3. Everyone connects to it
4. It works fine for 3 people and falls apart at 10

The ISP-provided router is designed to be “good enough for a house.” A shop, office, or warehouse is not a house. The walls are different, the area is bigger, the number of devices is higher, and the expectations are different.

The Basics: What You Actually Need

1. Separate your networks

At minimum, you need two Wi-Fi networks:

• Staff network — for your computers, POS systems, printers
• Guest network — for customers and visitors

Why? Because every device on your network is a potential problem. A customer’s phone with malware, a visitor’s laptop doing updates in the background — these shouldn’t be on the same network as your POS terminal.

Most decent routers support multiple SSIDs (network names). Set this up. It takes 10 minutes.

2. Put the router somewhere sensible

The ISP router usually ends up in the corner of the building where the phone/fibre line comes in. That’s often the worst possible location for Wi-Fi coverage.

Wi-Fi signals spread out in a sphere from the router. If the router’s in the corner, half the signal is going outside your building. Ideally, the router (or access point) should be centrally located, elevated, and not hidden in a metal cabinet.

If you can’t move the router, that’s what access points are for (more on that below).

3. Use 5 GHz for business devices

Most modern routers are dual-band — they broadcast on 2.4 GHz and 5 GHz.

• 2.4 GHz: longer range, slower speeds, more interference (everything from microwaves to Bluetooth uses this band)
• 5 GHz: shorter range, faster speeds, less interference

For your business devices — POS terminals, staff computers, printers — use 5 GHz. It’s faster and more reliable. Reserve 2.4 GHz for devices that are further away or don’t need the speed.

4. Get the right equipment

If you’ve got a small office (under 150 square metres, under 15 devices), a good consumer router might actually be fine. Look for:

• Wi-Fi 6 (802.11ax) — current standard, better with multiple devices
• Multiple SSIDs (for staff/guest separation)
• Gigabit Ethernet ports (for wired devices)

Brands like TP-Link (Deco series), ASUS (ZenWiFi), or Netgear (Orbi) have solid options in the $200-$400 range.

If you’ve got a larger space, multiple floors, or more than 20 devices, you need a proper access point setup. This is where it gets more serious:

• Ubiquiti UniFi — the go-to for small business. A UniFi access point ($150-$250 each) managed by free controller software. One AP covers about 100-150 square metres. Add more as needed.
• TP-Link Omada — similar concept to Ubiquiti, slightly cheaper, also good.
• MikroTik — powerful but steeper learning curve. Best if you’ve got some networking knowledge.

5. Wire what you can

Wi-Fi is convenient, but wired is always better for reliability. If a device doesn’t move — a desktop PC, a POS terminal, a printer — run an Ethernet cable to it. It’s faster, more reliable, and it takes load off the Wi-Fi for devices that actually need wireless.

Yes, running cables is a pain. But for fixed devices, it’s a one-time job that pays off forever.

How to Do a Basic Site Survey

Before spending money on equipment, understand what you’re working with:

1. Walk the space with your phone. Use a free Wi-Fi analyzer app (like WiFi Analyzer on Android or Airport Utility on iOS).
2. Check signal strength in every area where you need Wi-Fi. You want at least -65 dBm (the number is negative — closer to 0 is stronger). Below -75 dBm and you’ll have problems.
3. Note dead zones — areas where the signal drops off completely.
4. Check for interference — how many other Wi-Fi networks are visible? If you’re in a busy area (strip of shops, office building), the 2.4 GHz band is probably congested. That’s another reason to use 5 GHz.

This takes 15 minutes and tells you exactly what you’re dealing with.

Common Mistakes

• Too many devices on one router. ISP routers are not built for 30+ devices. If you’ve got a busy shop, they’ll struggle.
• Wi-Fi extenders. These are usually a bad idea. They cut your speed in half and create a separate network your devices have to switch between. A proper access point on a wired backhaul is always better.
• Ignoring the guest network. Customers on your business network is a security risk and a bandwidth hog.
• Never changing the default password. If your Wi-Fi still has the password from the sticker on the router, fix that today.
• Forgetting about uploads. Most internet plans are asymmetric — fast download, slow upload. If you’re doing video calls, cloud backups, or VoIP, upload speed matters. Check your plan.

The Bottom Line

Reliable Wi-Fi isn’t about spending the most money. It’s about understanding what you need, putting the right equipment in the right place, and separating your networks properly.

Start with the basics: separate staff and guest networks, use 5 GHz for business devices, wire what you can, and if the ISP router isn’t cutting it, invest in a proper access point. For most small businesses, $300-$500 in equipment solves 90% of Wi-Fi problems.

I’ve put together a site survey guide and router configuration walkthrough on Patreon — including a site survey template, recommended settings for common router brands, and a step-by-step guide for setting up separate staff and guest networks. Get it here.

When I ask, I usually get: “It runs our accounting software” or “That’s where our files are.” Sometimes it’s: “Honestly, I’m not sure what it does. Bob set it up before he left.”

Let’s fix that. If you’re paying for a server you don’t need, that’s money wasted. And if you’re not using a server when you should be, that’s a risk. Here’s how to tell the difference.

What a Server Actually Does

First, let’s clear something up. A server is just a computer that provides a service to other computers on the network. That’s it. It doesn’t have to be a $5,000 rack-mounted Dell. It could be a mini PC running Linux. The concept is what matters.

Servers typically handle:

• File storage and sharing — everyone’s documents in one place, with permissions
• Centralised backups — all machines back up to one location
• Running shared applications — accounting software, databases, line-of-business apps
• User management — Active Directory for login credentials and policies
• Email — if you run your own mail server (rare for small business)
• Print sharing — managing printers across the network

When the Cloud Replaces the Server

For a lot of small businesses, the cloud already does most of this. Here’s the mapping:

If your business runs on M365 and Xero, and your files are in OneDrive, you might not need a server at all. Seriously.

When You Still Need a Server

There are legitimate reasons to keep a server on-premises:

1. Line-of-business software that requires it. Some older (or niche) applications need to run on a local server. If your industry-specific software requires a Windows Server backend, that’s not something you can just cloud away.

2. Internet reliability. If your business is in an area with unreliable internet, relying entirely on the cloud is a gamble. A local server keeps things running when the connection drops.

3. Large file workloads. If you’re working with large files — video editing, CAD, large databases — transferring everything to the cloud and back is slow and expensive. Local storage is faster and doesn’t eat your bandwidth.

4. Data sovereignty or compliance. Some industries have requirements about where data physically lives. If you can’t put data in the cloud due to regulatory requirements, on-premises is your only option.

5. Latency-sensitive applications. If an application needs ultra-low latency to a database or file server, a local machine will always beat a cloud connection.

The Hidden Cost of “Free” Servers

That old server under the desk isn’t free. Here’s what it’s actually costing you:

• Power: A server running 24/7 draws 100-400W depending on load. At NZ electricity rates, that’s roughly $500-$1,600/year.
• Hardware replacement: Servers last 5-7 years. When they die, a replacement is $2,000-$5,000.
• Maintenance: Updates, patches, troubleshooting. If you don’t have in-house IT, that’s an MSP callout every time something goes wrong.
• Backup: You need to back up the server itself. That’s another cost.
• Security: An unpatched server on your network is a liability. It needs monitoring, firewall rules, and regular updates.
• Noise and space: It’s under someone’s desk. It’s hot. It’s loud. That’s not nothing.

Add it up and a “free” server is costing you $1,000-$3,000 per year, minimum.

The Cloud Isn’t Free Either

On the flip side, cloud services have ongoing costs:

• M365 Business Premium: ~NZ$36/user/month (excl. GST, annual billing)
• Cloud backup: ~$5-$10/device/month
• Cloud accounting (Xero): ~NZ$35-$125/month (excl. GST)
• Cloud storage beyond included tiers: varies

For a 10-person business, you’re looking at roughly NZ$450-$600/month in cloud subscriptions. That’s NZ$5,400-$7,200/year.

Neither option is free. The question is which set of costs and tradeoffs makes sense for your business.

The Decision Framework

Ask yourself these questions:

1. What does our server actually do? If you can’t answer this, that’s your first problem.
2. Can each of those functions move to the cloud? List them out. Check.
3. What’s the total cost of the server (power + hardware amortised + maintenance + backup)?
4. What’s the total cost of the cloud alternative (subscriptions for each service)?
5. Do we have compliance or connectivity constraints that prevent cloud migration?
6. What happens if the internet goes down? Can the business function for a day without cloud access?

If the cloud alternative is cheaper, your internet is reliable, and nothing requires on-premises — decommission the server. Redirect that money and attention elsewhere.

If you’ve got a genuine need for a local server, keep it. But maintain it properly. A neglected server is worse than no server at all.

The Bottom Line

The default assumption in 2026 should be cloud-first for small business. If you can do it in the cloud, do it in the cloud. Only run a server on-premises if you have a specific, identified need that the cloud can’t meet.

And if you do run a server, know what it does, keep it updated, and budget for its replacement before it dies at the worst possible time.

I’ve put together a full Server vs Cloud decision guide on Patreon — including a cost comparison template you can fill in with your actual numbers, a migration checklist for moving services to the cloud, and a “decommissioning plan” for when you’re ready to turn that dusty tower off for good. Get it here.

What to expect

• Troubleshooting guides
• Automation scripts and tips
• Infrastructure deep dives
• Backups
• General IT guidance gold!

Stay tuned for more!