Fast answer
The most common causes we find, in order: channel interference from neighbouring networks, a single consumer-grade access point carrying too many devices, DHCP pool exhaustion, failing PoE injectors or switch ports, roaming misconfiguration between multiple APs, ISP modem issues masquerading as WiFi problems, and RF-hostile construction. Each has a specific, verifiable fix.
1. Channel interference from your neighbours
In a multi-tenant building, dozens of networks compete for the same 2.4 GHz channels. When two access points transmit on overlapping channels, both networks slow down and drop connections intermittently - usually at busy times of day, which is why the problem seems random.
The fix: a spectrum scan to find clear channels, moving your business traffic to 5 GHz or 6 GHz where far more channels exist, and setting channels statically rather than trusting auto-selection, which often makes poor choices in dense buildings.
2. One consumer access point doing an enterprise job
A consumer router is designed for a family, not 25 staff with laptops, phones, VoIP handsets, and printers. Past 15-20 active devices, cheap hardware silently drops clients to cope. The device count is the problem, not the internet plan.
The fix: enterprise access points rated for high client density, placed according to an RF site survey rather than wherever a cable happened to reach.
3. DHCP pool exhaustion
Every device needs an IP address from a shared pool - many small-office routers default to ~100 addresses with long lease times. Guests, phones, and smart TVs quietly fill the pool, and new devices can no longer connect while existing ones drop when leases renew. This one produces the classic "WiFi is connected but nothing loads" complaint.
The fix: right-size the DHCP scope, shorten guest lease times, and separate guest devices onto their own network segment so they never compete with business devices for addresses.
4. Failing PoE, switch ports, and cabling
- Access points powered over Ethernet reboot silently when a PoE injector or switch port starts failing - users just see WiFi vanish for 90 seconds.
- Cheap or damaged cable runs cause packet loss that looks exactly like a wireless problem.
- Cable terminated years ago by whoever was handy rarely passes a certification test.
The fix: test the wired path first. We cable-certify runs and check switch logs before touching wireless settings - about a third of "WiFi problems" we investigate are actually wired-side failures.
5. Roaming misconfiguration between access points
With multiple APs, your laptop decides when to hop between them - and clients are notoriously sticky. They cling to a distant AP at one bar of signal instead of switching to the nearby one. Calls drop as people walk to the meeting room.
The fix: enterprise wireless supports 802.11k/v/r standards that guide clients to roam at the right moment, plus minimum-signal thresholds that gently push sticky clients off overloaded or distant APs.
6. It's not the WiFi - it's the ISP or the firewall
If wired desktops stutter at the same moments the WiFi "drops," the wireless is innocent. Failing cable modems, saturated uplinks, and undersized firewalls all present as flaky WiFi because most users only experience the network wirelessly.
The fix: monitoring that records latency and packet loss on the WAN link separately from the wireless layer, so the diagnosis is data, not guesswork. This is standard in our managed IT plans.
7. The building itself
Concrete, brick, metal ceilings, glass partitions, and even large aquariums attenuate WiFi differently. New Westminster heritage brick and warehouse steel racking are the two most hostile environments we survey in the Lower Mainland.
The fix: a heat-mapped site survey that measures your actual building, then AP placement and antenna selection engineered for it. Guessing is how offices end up with four extenders and worse WiFi than they started with.