Here's how you can use technology and tools to evaluate performance in Internet Services.

Quantitative is a start, but you will also need qualitative monitoring tools to establish how much of it is waste traffic with little or no potential value.

To ensure your Internet services run smoothly, use tools like network monitors to look for issues, performance testers to check how fast things load, and user experience trackers to see how people interact with your site. Tools like Speedtest help you measure speed and latency, while security scanners like Nessus find potential risks.

If you can’t measure it, you can’t manage it. Remember, KPI stands for "KEY" Performance Indicators, not all performance indicators. It’s easy to get bogged down by metrics that don’t matter. I focus on crucial metrics like uptime, load time, and capacity because our services demand reliability. Our BGP peers with Spectrum and Lumen need over 99.997% uptime. Starlink, used for out-of-band access, doesn’t need to be as robust but still requires monitoring. I used to track every metric. Now, we focus on fewer, meaningful KPIs. We review key metrics weekly, while our monitoring system handles less critical issues. For specific customer needs we may change the scope for monitoring and KPI development.

Google and Facebook are excellent resources for gathering customer feedback. At Hostirian, we emphasize online feedback and social scoring. One key metric I focus on is the Net Promoter Score (NPS). I recommend asking your clients annually, "On a scale of 1 to 10, how likely are you to refer business to us?" This simple question can offer profound insights into customer satisfaction and loyalty.

Auvik is excellent for monitoring, but it's not the only tool. We also use custom scripts to purge logs and alert on specific details. Monitoring for DDoS attacks and changes in data center temperature is vital. You're not just watching for failures but for significant changes that could indicate problems. Treat your monitoring system as a living entity, constantly evolving. Predefined thresholds are starting points, but ongoing adjustments are essential for effective network and system monitoring.

At Rivercity Internet Group, we follow a SMaC (Systematic, Methodical, and Consistent) rule set. One fundamental rule is: if it can be automated, it will be automated. However, don’t fall into the trap of forgetting about automation once it's set. Everything automated needs monitoring to ensure it runs smoothly and completes its tasks. I’ve seen scripts run fine for years, only to fail when an update breaks the automation. Regular checks are essential to maintain effectiveness.