In our previous blog, we discussed starting your collaboration network analysis with basic call trends. Do you know how many calls are processed in your network each hour? Any idea how many calls complete normally versus fail each day? How many calls dropped yesterday? Do you know the busy days or what traffic load your enterprise has in off hours? This blog will highlight some examples of basic call trend analytics and the vital information you can discover.

In general, if you think about what you expect a graph of the hourly call counts in your enterprise to look like, it might be something like the example below. This shows hourly call trends for five days, starting on Thursday and running through Monday afternoon. Typical pattern for many customers with call counts building in the mornings, maybe a dip in the lunch hour and then dropping off to small numbers overnight and on weekends.

However, we know that there are not often “typical” patterns in our world of collaboration network management! Below are snapshots from three real customer networks. Each shows the hourly call trends for five days. For each customer, there are two different charts presented with the first showing a Friday through Tuesday afternoon in December and the second representing a Thursday through Monday afternoon in January. Do you know which one might look most like your enterprise?

For the first customer, you will notice a pattern of around 400 calls/hour during business hours with very little weekend traffic for the week in December. The call volumes increased to approximately 500 – 600 calls/hour for days in the January chart. Back to work after the holidays!

These charts not only show the normal calls that were setup and completed as expected, but they also highlight how other calls failed to setup initially or were terminated unexpectedly. Examples include barred by policy, dropped calls or calls that were “threats,” basically representing inbound call hacking attempts. For this customer, the light purple line clearly shows spikes of calls throughout the day for both weeks for “Party Not Available” cases. Is it OK that there are that many calls not being answered? And why would there be that huge spike on Saturday night, January 6?

Customer two illustrates a completely different trend. Note that there are dips in traffic in overnight hours but clearly lots of weekend traffic. This customer typically has a peak call rate around 600 calls/hour. Note this customer has very few failed or non-normal calls. A very “clean” network until that huge spike of “Party Not Available” calls on Friday, January 5. With the ability to monitor this trend, this customer was quickly able to identify the source of the problem (a routing error) and address it.

The final customer example again shows more traditional weekday call trends with low traffic on weekend days. This customer traditionally has less than 150 calls/hour but there are three very notable spikes of threats in the December week highlighted by the orange line. These are calls that are inbound to the network with a call resolution of “Called Number Failure” based on the clearing cause code assigned by the switch. In most cases, these represent inbound call hacking attempts. Is your network experiencing any spikes like this? There is also a spike of over 250 calls in an hour that were not answered on December 18. For this customer those hacking attempts and unanswered calls continued to spike as the new year started.

So, how does your network compare? How many calls on your network are failing to even setup? Is it a small percentage of your total calls or not? Are calls dropping throughout the day? Are you using subjective measurements like trouble tickets created to track issues? The right forensic tools will give you visibility into usage trends but will also speed trouble identification and resolution throughout your network. Come back soon for the next post to dig in deeper to analytics for transforming your network management and capacity planning process.

Clearly See Your Call Trends and Performance with Optic
Optic for Voice Networks™ will give you a clear view into your network’s usage, trending and performance, both from high level trending dashboards to the ability to search for specific calls and see detailed metrics. Visibility will help you improve caller experience and streamline system administration. Contact us today to learn more about proactively monitoring your voice network and gaining seamless visibility across your entire voice network activity and performance.