The WiFi network isn’t just another modern perk. It is the backbone of business operations around the world and it can’t be taken for granted. These networks aren’t a static resource, but a dynamic force, one that can be better or worse depending on numerous factors. Since no organization wants “or worse” to be the status of their network, how do you keep your network performing optimally? Here is a look at key performance indicators (KPIs) and end user quality metrics that everyone should be monitoring.

Contents

How is WiFi performance measured?

WiFi end user quality metrics

How to capture end user quality metrics

Interpreting and improving end user quality metrics

Additional resources

How is WiFi performance measured?

There are many metrics monitored and analyzed by IT professionals. However, we think the best way to measure WiFi performance is with end user quality metrics. These metrics measure how well the network performs for end users – i.e. employees, customers, executives, etc. If the network isn’t performing to a standard that meets user expectations and needs, then it isn’t performing well.

End user quality metrics tell IT professionals if the WiFi network is secure, fast, available, and future-proofed. These metrics identify factors that have a direct impact on users, and include these WiFi network KPIs:

• Uptime
• Signal strength
• Noise level
• Signal to noise ratio (SNR)
• Packet loss
• Retransmission
• Latency
• Bandwidth and throughput
• Network jitter

They also require IT to have in-depth knowledge of historical WiFi performance and complete network visibility in order to identify network interference and congestion.

WiFi end user quality metrics

Here is a breakdown of these important metrics:

1. Uptime – aka the gold standard, influenced by all other metrics. In short, uptime is the amount of time your WiFi service is available. Uptime should be 24/7. If you’re having any issues with a lack of uptime, then you have downtime, and you’ve got problems.
2. Signal strength – this KPI measures how strong your WiFi signal is throughout your workspace. If you want end users to have high quality WiFi, then signal strength needs to be pervasive. The optimal signal strength will differ from space to space, and not every space needs the same strength. If there are fewer applications in use and minimal interference, a lower signal strength can still deliver optimal performance. However, most organizations are dealing with hundreds to thousands of devices and a fair amount of interference. This makes signal strength key to understanding if your end users can perform their jobs or not.

Signal strength is measured in two ways: RSSI (received signal strength indicator) and decibel milliwatts (dBM). RSSI differs from vendor to vendor, but the higher the number the better. dBM, on the other hand, is always negative. -20dBm is a higher signal than -50dBm.

1. Noise level – noise is any signal interference that doesn’t come from WiFi sources, such as interference from cordless phones, radar, microwaves, refrigerators, or TVs., building materials, or large bodies of water. The noise level is measured in negative decibels from 0 to -120. The closer you are to -120, the better. The closer you are to 0, the more you will experience data corruption, retransmissions, and degrading throughput and latency. 
2. Signal to noise ratio (SNR) – SNR measures how much WiFi signal is actually available compared to all the other signals (noise) that can interfere. If an organization shows great signal strength, but poor network performance, it’s time to check SNR. You do this by measuring signal and noise, and then subtracting noise from the signal strength. The higher your value, the clearer your signal.
3. Packet loss – every action and transmission on a WiFi network uses data packets. Some measure of packet loss is expected. Too much however will affect application performance and is a sign of interference, congestion, and/or low bandwidth. Some applications, like video conferencing tools, will be ruined by 1-2% packet loss. Other less data-intensive apps can survive with packet loss up to 3%.
4. Retransmission – retransmissions occur whenever the original data transmission didn’t work. If packets are continuously lost, and continuously retransmitted, this reduces the overall throughput available to a business. Every user will experience network and application delays.
5. Latency – this KPI is a measure of how long it takes data packets to travel from one point to another. High latency results in issues such as long buffering times for videos, long transmission times for emails, and slow loading applications and web pages. 50 milliseconds is a general baseline for maximum allowed latency for high-performing applications (the kind that you want your end users to have).
6. Bandwidth and throughput – these two metrics might be the ones most familiar to non-IT professionals. Bandwidth is how much data a WiFi network can optimally transfer and throughput is how much is actually being transmitted during a certain period of time. Throughput should be continuously monitored so that IT can ensure the network is delivering at its full potential.
7. Network jitter – a jittery network is one that is inconsistent with its data transfer rate. It will often delay sending packets, which causes high latency. For this reason, jittery networks are usually associated with video performance problems. Jitter can be caused by network congestion, physical interference, and interference from radio sources. It is this last factor that makes WiFi networks more prone to jitter than wired networks.
8. Historical analytics – while not a commonly cited end user quality metric or KPI, historical analytics are incredibly important for any organization wanting to streamline network performance and operational efficiency. These analytics reveal long-term performance and behavior trends; trends that might not be identifiable in real-time because they still fall within established performance ranges, but nevertheless reveal a degrading network. 
9. Complete network visibility – this isn’t a metric, per se, but it is required if any organization wants to confidently state that it has true insight into the end user experience and network behavior. To be complete, visibility must be 24/7, use real-time and historical analytics, and extend into all aspects of the network environment: connected devices, backend and frontend infrastructure, and external sources of interference such as nearby networks and Bluetooth devices.

With these metrics, IT can build a comprehensive picture of a network’s health.

How to capture end user quality metrics

The name of the game is actionable knowledge. WiFi networks consist of hundreds or thousands of devices, with tens of thousands of data packets sent every second. They are dynamic, constantly impacted by the number of devices on the network, device utilization, workspace layout, the location of devices, and nearby networks.

If relevant, usable metrics are to be captured to assist IT in optimizing such an environment, these metrics must be captured constantly and analyzed in real-time. This necessitates the use of automated tools and solutions.

Automated solutions can work nonstop. They don’t require lunch breaks and have no end to their work day. They are always onsite and can analyze packets from every device in seconds. They provide IT with the information that professionals need to understand and optimize the network, improving user experience and boosting operational efficiency.

There are a number of automated tools for performance testing. Every organization needs to find the one that best fits their needs. This will be based on number and type of locations, users, devices, and network utilization.

With that said, there are a few key elements that any automated solution should provide.

• AI capabilities
• Root cause identification
• Historical and real-time analytics
• Proactive alerts
• Remote troubleshooting

AI capabilities

In the world of network optimization, AI solutions ease the workload of IT professionals and simplify problem identification and mitigation. Solutions powered by AI can learn to recognize normal and abnormal network behavior, alert IT when things seem to be going wrong, identify the root cause of issues, and track network performance over time. Ultimately, every other required element on our list is made possible by AI.

Root cause identification

IT can’t solve problems until they know what is causing the problem. This requires root cause identification. If a solution doesn’t automatically provide the root cause, professionals have to identify it themselves. This requires them to see the problem in action, capture packet data at that moment, and then sift through and analyze the data on their own time. In other words, this manual process is slow and energy consuming. Solutions that deliver the information automatically save companies a lot of time and money.

Historical and real-time analytics

Real-time analytics allow IT to resolve problems as soon as they occur, often before users are ever impacted. Historical analytics let decision makers see whether or not, and how, networks are degrading so that cost-effective upgrades can be implemented before issues start. Both types of analytics are key to an optimized, future-proofed user experience.

Proactive alerts

IT should be notified as soon as an issue occurs or a test’s results fall outside of established parameters. This proactive alert allows IT to resolve the issue as soon as possible, rather than waiting for a “downstream” notification from an unhappy user.

Remote troubleshooting

No matter how many metrics IT analyzes, and how quick teams can resolve problems, there is always the chance another issue is going to crop up. Issues can occur at any time and any location – making troubleshooting especially difficult for IT teams responsible for multiple spaces. If teams are supported with remote troubleshooting, they can review end user quality metrics and implement any needed resolutions from anywhere.

Interpreting and improving end user quality metrics

While the right automated solution should significantly decrease the number of performance problems, every issue that does occur tells IT something about the network.

• Is performance degrading?
• Is utilization changing?
• Is a specific device or application not performing as promised?

If an issue keeps recurring, that is a sign that something needs to change. If it’s a signal problem, perhaps the network layout isn’t optimal. If it’s an application problem, maybe changes need to be discussed with the vendor. Whatever the case, every issue should be studied in two ways:

1. In isolation – i.e. fix the problem that needs to be fixed
2. Comprehensively – use the problem to form a complete picture of network performance and understand exactly what end users experience

This comparative analysis and interpretation will help you improve metrics and your WiFi network.

Additional resources

Keeping a business running seamlessly is a superhuman task, a responsibility that IT professionals must accept every day. Giving these professionals the support and resources they need will improve business for everyone. Work with an AI-powered WiFi automation solution and build a workplace that guarantees a positive brand image and end user satisfaction.