As with most companies, we keep track of a few metrics to help us drive improvements and ensure that we focus our efforts on the right things.
One of the metrics that is very near and dear to our hearts is the reliability of our products. We use BigQuery to store analytics data, and Klipfolio to turn that data into pretty dashboards like this one:
The above charts are a snapshot of what usage and reliability look like for Rise Cache (the software that keeps your bandwidth costs to a minimum by caching content) and a few of our Widgets. The green lines show usage while the blue lines plot reliability. The percent value indicates today's reliability. You've probably already figured it out by now, but red = bad and black = good.
While we always strive to achieve a reliability of 99% or better, occasionally we do drop below that threshold. When that happens, we typically analyze our BigQuery data more closely to try to find the cause of the decrease, and then make any adjustments as necessary. We also want to be sure that the drop in reliability is not a symptom of a more serious problem.
How We Measure Reliability
Let's use the Image Widget as an example. When the image is shown on a Display, or if an error should occur (such as if the image fails to load), the details are logged to BigQuery. We then use that usage and error data to calculate the reliability:
Reliability = (# of Displays using Image Widget - # of Displays with Image Widget errors) / # of Displays using Image Widget
We calculate the reliability for each of the last 30 days, and display the result in nifty charts like the ones you saw earlier.
We hope that you can rest assured in the knowledge that we take the reliability of our software very seriously. It's become a part of our daily routine to inspect these metrics, plus others, to ensure that your digital signage continues to run smoothly day in and day out. If you have any questions on our reliability or anything else, leave us a comment, we love hearing from you!