Treating Messaging as Critical

Efficient communication is an important element of daily work. In addition to this intrinsic need, regulatory pressures, the need for security and rising volumes of both email and instant messaging are taxing IT organizations like never before.

Why then, isn’t more attention paid toward treating the various messaging platforms as critical systems in organizations? If you haven’t looked carefully at your messaging infrastructure, now is the time.

Conduct a Risk Assessment

First, if you haven’t done so already, conduct a risk assessment of services provided by IT. Ensure that email is included in all systems/services lists so the various people involved do not overlook the ubiquitous email. It is easy to think of the financials, manufacturing and so on, but email and IM are increasingly being taken for granted as a “given” infrastructure element. In most organizations, messaging, be it email and/or IM, are critical and must be increasingly managed with the same care traditionally afforded the ERP and other critical systems.

Metrics

For messaging systems, it is vital to have meaningful, accurate and timely statistics that show utilization, potential and real problem areas, etc. All critical systems must evolve to the point that metrics can be established and ingrained in management decision-making — messaging is no different.

Messaging needs the same care in crafting, collecting and analyzing performance metrics as any other critical system. To meet these needs, there are vendors who have studied the needs of organizations and developed the necessary tool sets.

One of the leading vendors of application performance management solutions specifically for messaging and collaboration is DYS Analytics of Wellesley, Mass. Its DYS CONTROL! application suite is in use by numerous customers, including JP Morgan Chase, ABN AMRO, IBM, PWC and Zurich Financial Services.

A Hierarchy of Needs

One interesting thing about DYS is its adoption of Maslow’s Hierarchy of Needs diagram to messaging to provide a framework as to what metrics are needed at what stages of messaging’s adoption in the organization:

They work with their clients to use metrics collected by the DYS CONTROL! application suite to move up the messaging hierarchy of needs. This happens by using metrics to make accurate decisions about management of the messaging platforms, the underlying network, service levels, problem users, and so on. Questions need to be asked at each level and needs met before the organization can proceed to the next level.

What Metrics?

Of course, the above discussion spawns the need to identify what some of the metrics are that an organization needs to track. While each organization needs to develop a list based on their environment, a short general list includes:

Delivery Time — how long it takes a message to go from one user to another. This helps establish if service levels are being met. This information needs to be at the user, department and organization levels, or however service level agreements are set in the organization.

Storage Utilization — The amount of mail database usage needs to be tracked at the user, department and organizational levels. This information helps identify abnormally high users for investigation, set service levels, establish charge back allocations, plan for growth accordingly and so on.

Messaging Utilization — how many messages and how many megabytes of data is each user transferring? Again, this should be rolled up to department and organization levels. The same recurring themes exist — identify heavy users, look for problems, identify trends, and plan accordingly.

Server Activity — what is going on with the server? This may range from current routing activity to the queue time. The intent is to have visibility to real time as well as historical activity.

Top Destination Domains — A list of the top destination Internet domains. For example, an organization may find that the top destination domain is aol.com for 10,000 messages and 1GB of data. The question certainly arises as to why this is the case. It may be that power users are emailing themselves work to do at home and another transport method is needed.

The metrics that you need depend on the size of your organization, messaging systems used, goals of the organization, goals of IT, service level expectations and so on.

In addition to collecting the metrics, there must be a formal process to routinely review the metrics and take appropriate action. All to often, technology is purchased and then ineffectively used. Regardless of how the metrics are collected, there must be a defined process to use the data — otherwise it is not worth the time and costs to collect the metrics!

Without a doubt, messaging systems are critical in this day and age. Organizations and their IT departments must recognize this need and act accordingly. Metrics must be collected in order to assist in problem resolution, capacity planning, charge backs, etc. It is vital that organizations adopt tools that assist in meeting the needs of their stakeholders by automating the collection of metrics data and aiding in subsequent analysis to ensure appropriate attention is given to the critical messaging systems.