If you run five, fifty, or five hundred locations, you can't walk every building to see whether the network is working. You need a short list of numbers that tell you, at a glance, which sites are healthy and which aren't — ideally before staff or customers start calling.
These are the seven we recommend every multi-site operator track.
1. Access point and site availability
What it is: the percentage of time each access point, switch, and site uplink is online.
Why it matters: it's the foundation. An offline access point in a lobby is an inconvenience; one in a care wing, a stockroom, or a kitchen can stop work. Track it per site and per device, and alert on outages in real time — not in a weekly report.
2. Client experience (connection success)
What it is: the share of connection attempts that succeed, and how long it takes devices to associate, authenticate, and get an IP address.
Why it matters: a network can be “up” and still fail users. Authentication errors, DHCP exhaustion, and DNS problems all look like “the Wi-Fi is broken” to the person holding the device.
3. Coverage quality
What it is: the percentage of connected devices with healthy signal strength and signal-to-noise ratio.
Why it matters: a rising share of weak-signal clients in a particular area is an early warning of a coverage gap, a failed access point, or a physical change such as new shelving or a renovation.
4. Capacity and utilization
What it is: channel utilization, clients per access point, and throughput at peak times.
Why it matters: networks fail at the busiest moment — shift change, lunch rush, a conference session, a holiday weekend. Peak-hour data tells you where you need more capacity before it becomes a complaint.
5. Roaming performance
What it is: how often devices move between access points, how long handoffs take, and how often they fail.
Why it matters: for mobile staff — caregivers, pickers, servers, security — roaming is the experience. Slow or failed roams mean dropped calls and frozen apps.
6. WAN and internet health
What it is: latency, packet loss, and availability of each site's internet circuit and failover link.
Why it matters: most of what users do lives in the cloud. A perfect local Wi-Fi network is useless behind a saturated or flapping circuit — and failover links that are never tested tend not to work when needed.
7. Time to detect and time to resolve
What it is: how long it takes for your team to learn about a problem, and how long it takes to fix it.
Why it matters: this is the KPI that tells you whether your monitoring actually works. If most issues are discovered by a phone call from the site, time to detect is your biggest opportunity.
Tip: compare sites against each other, not just against a fixed threshold. The location performing two standard deviations worse than its peers is usually the one that needs attention, even if it's technically “green.”
Turning KPIs into operations
The real value comes when network data connects to how the business runs:
- Occupancy and foot traffic from device counts can inform staffing and space planning.
- Dwell time and visit patterns help retail and hospitality teams understand customer behavior.
- Device presence can confirm that critical equipment is online and where it should be.
The privacy rule of thumb: use aggregated, anonymized data for operational insight, and be transparent with users about what's collected.
Common pitfalls
- Too many dashboards. Different vendors at different sites means different tools. Normalize data into one view.
- Alert fatigue. If everything alerts, nothing does. Alert on what needs action and route it to the right person.
- Averages that hide problems. A healthy portfolio average can conceal one site that's failing every afternoon.
The bottom line
You don't need hundreds of metrics. You need seven good ones, collected consistently across every site, with alerts that reach someone who can act. That's the difference between a network your team reacts to and one they actually manage.