In my last two columns, I addressed two important aspects of business intelligence (BI) governance. I continue with a third aspect - total cost of ownership (TCO) management. TCO is a term commonly used to identify all of the expenses associated with developing and operating the BI program and its components. TCO provides valuable information that can be used to determine the return on investment (ROI) of the program.

The traditional TCO approach addresses the visible expenses associated with BI, which means that it normally focuses on the hardware, software, telecommunications, and development and support personnel costs. This approach leaves out an extremely important component, and several clients have struggled to justify moving to an enterprise data warehouse because that component was omitted. The missing aspect is the time (and hence, cost) of the business analysts who use the information. I propose that instead of the traditional approach, companies should emphasize TCO management of information delivery. This subtle change in terminology expands the scope from the technical environment to the business environment and implies that - through the inclusion of the term management - action will be taken based on the cost information that is captured.

To understand the impact of including the business analyst's time, consider a company trying to justify the migration from a data mart silo environment to an enterprise data warehouse or to the development of conformed dimensions. If it calculates the TCO using the traditional approach, the enterprise data warehouse and conformed dimensions may (at least in the short run) appear to be more costly. Without exceptionally strong executive support, such a project would not be perceived as being cost justified. With the proposed approach, the time business analysts spend capturing, validating, reconciling and correcting data should be included. It is not uncommon for these analysts to spend a major portion of their time on such activities, leaving only a small portion of their time to actually use the information. The enterprise approach addresses the information capture and reconciliation (by consolidating it into one source) and the information validation and correction (by addressing data quality in the extract, transform and load  processes). A company with 100 analysts and an average weighted annual cost of $100,000 per analyst has an expense item of $10 million per year for salaries. If their productivity is improved by as little as 10 percent, the TCO of the silo approach will reflect $1 million in costs that the enterprise approach will not. This provides a much better comparison base and should help justify the migration to the enterprise approach.

Measuring all the components of TCO is critical if we are to manage it. Ideally, we would capture the actual time spent by business analysts on information delivery activities, and with that information, we would have our cost components. In many companies, this is not practical. Therefore, we need another approach for capturing these costs. My suggestion is to accomplish this through a periodic survey, which includes questions such as:

  • What percentage of your time do you spend searching for information? What particular data is difficult for you to find?
  • How many sources of information do you use to capture information for your most common analysis and reporting activities? What data has the most sources, and what are the sources you use?
  • What percentage of your time do you spend reconciling data obtained from multiple sources? What data is most difficult or time-consuming to reconcile?
  • What percentage of your time is spent reworking analysis because of data errors that you or others discover? What data is most prone to such errors?
  • What percentage of your information analysis and delivery time do you feel could be saved if accurate, reconciled, trustworthy information were available from a single source? What improvement (if any) have you experienced since the last survey?

The first survey should be conducted as early in the BI program as possible, with subsequent surveys performed following incremental deployments. Primary and secondary sets of questions are included. The primary questions are designed to provide the estimated cost information for inclusion in the TCO calculation. While not exact, the information is revealing, and it becomes meaningful if it is tracked over time or used as the basis for projecting savings. The answers to the secondary questions indicate specific areas that are troublesome. This is useful for establishing BI priorities and may actually identify some short-term measures that can provide dramatic improvements. In addition, the improvements in specific areas will become more visible as the troublesome areas shift over time.
TCO management provides one set of measures that help understand the value of BI. Other success measures - both objective and subjective - should also be tracked.1 Some success measures to consider include usage levels, user satisfaction, visible data quality issues, reuse levels and, ultimately, the business value obtained.

Managing the TCO for information delivery is a critical success factor for a BI program. The included costs need to go beyond the traditional visible expenses and must address the effort people spend converting data they obtain into useful information that can be applied to provide business value. 

Reference:

  1. For the subjective metrics, the trend resulting from periodic measurements is often more significant than the actual value.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access