IT Management Reference Guide

This segment concludes the four-part case study on evaluating recovery sites for business continuity. Part one explored qualitative comparisons among the 12 cities involved in this analysis. Parts two and three presented a quantitative analysis of natural and man-made disaster issues, respectively. This final installment describes seven valuable take-aways that managers and business continuity planners can apply when evaluating their own alternative recovery sites. I summarize these seven take-aways in Figure 1, and then describe each of them in more detail.

Figure 1 Take-Aways from the Case Study on Evaluating Recovery Sites

1. Decide on the scope and quantity of geographic alternatives – Scope refers to the geographic area being considered for possible recovery sites. Managers and business continuity planners should make this determination as one of their first decisions because their choice will impact other key decisions such as the number of alternative sites to evaluate and the types and number of selection criteria to use. The geographic area of possible recovery sites may vary from being across town, to intrastate, to nation-wide, or even international.

   I have known some companies to use a different building within a single campus setting as their recovery site. Other companies may use a recovery site located a few miles away from the campus of the primary data center. Still other firms elect to use a facility hundreds of miles apart from their main IT center. While there is no exact guideline as to the optimal distance a recovery facility should be from its primary site – there are simply too many variables that impact various environments differently – a good rule of thumb is 50 miles. An even better rule of thumb is to identify the common risks such natural disasters, power outages or terrorist attacks that may be prevalent in the environment of the original data center but not present at a given recovery site.

   The quantity of recovery sites one should evaluate usually correlates to the size of the geographic scope. In general, the larger the area the greater the number of potential sites. As previously mentioned, in this case study the geographic area was nation-wide and the number of cities evaluated was 12. Some companies that are world-wide may require fewer than 12 alternative sites if they have well-quipped recovery facilities in other countries.

2. Agree from where your base location will originate – Many companies have multiple data centers in multiple cities. In order to do a comparative analysis of alternate recovery sites, one of these data centers needs to be designated as the primary site. In the case study, the primary data center was located in Irvine, California, even though there were plans to share workloads with a future data center elsewhere.

3. Decide on your major categories of selection criteria – The next step in the process of evaluating potential recovery sites is to decide on several major categories of selection criteria from which to analytically compare their values to each other. Examples of these categories include weather, climate, natural disasters, man-made disasters, quality and availability of workforce, economic issues, political environment and social factors. These categories should have several common characteristics that will help in the analysis. These include categories that are relevant to the environment of the company being evaluated, quantifiable for numerical comparisons, and able to be broken down into smaller sub-categories.

4. Decide on your major sub-categories of selection criteria – Each of the major categories should be able to be broken down into sub-categories. The ability to have each of these sub-categories quantified is even more important for this set of selection criteria than for the major categories. This is because the sub-categories are what will be used for the definitive quantitative analysis on which the recommendation of a recovery site is significantly based.

5. Use reliable sources to validate and quantify the major categories and sub-categories – Planners can usually find a wealth of information from various sources to validate and quantify the major categories and sub-categories. These sources include websites, vendors, government agencies, consultants, and other subject matter experts.

6. Determine what, if any, weights to apply to your selection criteria – The decision of whether or not to apply weights to the selection criteria is a trade-off between complexity and simplicity, between extensiveness and expediency, and between thoroughness and summarization. Applying weights to some or all of the sub-categories adds precision, accuracy and differentiation to the analysis process. For example, in our case study my client briefly considered using weights for areas under the natural disaster category. We eventually decided not to use this method because the benefits seemed small when compared to the time and complexity it would add to the process. But other companies may feel differently about the value of weights, and should consider the pro's and con's of their use at the outset.

7. Identify any political or corporate culture issues influencing site decision – No matter how diligent your efforts in conducting a thorough, unbiased, and totally objective analysis, there is always the possibility that subjective influences may eventually override your final recommendation. This does not mean that the work performed in doing the quantitative and qualitative analysis is not worthwhile. The selecting and evaluating of valid criteria can in and of itself be valuable in terms of ranking the alternative locations in the event that the top-ranked selection is not initially chosen. Identifying political or cultural factors at the outset may lead to a clearer focus on what criteria are likely to carry the most weight in the final analysis.

This concludes the four-part series on the case study of how to evaluate disaster recovery sites for a company's use for business continuity. Not every company will use the same selection criteria because the size, complexity and priorities will vary widely from enterprise to enterprise. But the process described in these four installments can be applied to virtually any organization wanting to select the optimal recovery site for their particular environment.