A client called us convinced their WiFi needed a hardware upgrade. Utilization on some access points was spiking to nearly 99%, week after week, on a predictable schedule. The obvious read was “too many people, not enough network.”
That would have meant new access points, new cabling in some spots, a real project with a real price tag. Before recommending any of that, we went and looked at what was actually happening on the network during those spikes — and the fix ended up costing nothing.
A spike isn’t a diagnosis
“Utilization is high” tells you something is busy. It doesn’t tell you why, and those are two very different problems to fix. High utilization from too many devices trying to share too little airtime calls for more access points. High utilization from a misconfigured access point burning capacity inefficiently calls for a config change. Guess wrong and you spend real money solving the wrong problem, while the actual cause keeps running.
We pulled two things before touching any hardware: the exact timing pattern of the spikes, and the current radio configuration on the access points involved. The timing lined up almost exactly with a recurring building-wide event — a real, predictable jump in the number of people and devices on-site. That part of the story was legitimate. But when we checked the radio settings, two access points were running a wider channel width than the environment could actually support cleanly, and the network’s minimum data rate on the 2.4GHz band was set low enough that a single struggling device could hold up airtime for everyone else nearby.
Two real problems, layered on top of each other
The traffic spike was real and not something to “fix” — it’s just what happens when a building fills up. But it was exposing a configuration weak spot that had been quietly costing capacity every single day, not just on the busy ones. The wide channel width meant more interference from nearby networks sharing the same spectrum. The low minimum data rate meant the access point kept slowing down to accommodate distant or older devices instead of serving everyone at a reasonable speed.
Neither of those would show up if you only looked at “is utilization high.” Both showed up the moment we compared configuration against what the manufacturer actually recommends for a busy, crowded RF environment.
What actually fixed it
Two configuration changes, no new hardware: narrowed the channel width back down on the two overconfigured access points, and raised the 2.4GHz minimum data rate so the network stopped waiting on its slowest possible client. We confirmed the fix by comparing utilization at matching time-of-week windows before and after — not just “does it feel better,” but the same Sunday-morning peak, the same Wednesday-evening peak, measured against themselves. Peak utilization roughly halved across every access point we changed, with the same or higher number of connected devices during the comparison windows. The busy periods were still busy. They just weren’t choking the network anymore.
Before you approve a WiFi upgrade, ask this
- Has anyone actually compared your network’s channel width and data rate settings against what’s recommended for your environment, or are they still on defaults from install day?
- Is the “high utilization” happening at the same times every week — and if so, has anyone confirmed whether that’s real device load or a config issue that happens to be more visible when it’s busy?
- Would you know the difference between “we need more access points” and “our existing access points aren’t configured for the environment they’re actually in”?
A capacity problem and a configuration problem can look identical on a utilization graph. Only one of them needs a check to be written.