The old saying among analysts goes “If you can’t measure it, you can’t monitor it.” Measurement and monitoring of networks, applications, and services has many kinds of value. Doing it right can make a profound impact on the bottom line, too.

Companies and organizations buy bandwidth for Internet access and other communications based on their current consumption levels, and on estimated forecasts based on observed historical trends – if they know them.  There can be a tendency to overbuy bandwidth, perhaps in anticipation of as-yet-unrealized peak usage levels, or perhaps just to be sure not to top out when that next peak comes.

Monitoring of actual network use, along with careful trend analysis is the best way to establish meaningful peak and average usage scenarios. It’s important to be able to handle known peaks, and to anticipate the peaks that lie ahead. But those companies that have reliable, long-term observations of actual traffic and usage patterns can use that data to right-size their WAN connections, especially when it comes to buying bandwidth. This helps to offset the inevitable tendency to overbuy in anticipation of rising demand, and saves substantial amounts of cash by reducing overall bandwidth costs. Careful attention to average usage and peak levels can also optimize use of a failover or supplementary link, or alert staff when a backup link is being used full-time. Since backup links can be expensive and billed based on usage, the financial implications are obvious.

For example, the IT staff at TRA Medical Imaging observed that their average bandwidth consumption was substantially lower than the 100 Mbps they were paying for. TRA is a Tacoma-area-based medical imaging and information firm with multiple locations, and relationships with numerous area hospitals. They have high bandwidth needs so they can accommodate transfers of high-resolution medical scans such as PET scans and MRIs between their multiple clinics, and to doctor’s offices, other clinics, and hospitals. By using OmniCenter to observe that they were consuming only about 40 Mbps of the 100 Mbps they were paying for, the CIO was able to save substantially on their bandwidth expenditures without impacting service. Armed with the data it gleaned from its observations, TRA renegotiated its wide area network contract with its service provider to replace an MPLS link with a mesh network and layer 2 networks, which are more secure and more reliable, as well as less expensive. The vendor agreed with TRA’s assessment and its request for a service change, and ended up reducing monthly costs for the WAN by half. The overall result was better connectivity at a much nicer price.

Perhaps there’s an important corollary to the old saw we mentioned at the start of this post – namely, “If you don’t monitor it, you can’t manage it.” Fortunately for TRA, and for other organizations savvy and careful enough to match purchases to actual consumption levels, that also means you can sometimes save on outlays necessary to pay for network services by buying only as much as you need, in a form that provides the best combination of speed, function, and reliability. Beyond bandwidth management, such monitoring can also keep your IT staff informed about applications and services in use, and provide a basis to drive policy on unapproved use, and to make sure that the return on networking investments is maximized.