Data Center Management Highlights

Server upgrade strategy for a large datacenter in the internet information services industry.

Issues

Several months required for internal analysis.

Critical errors and lack of transparency in analyst's spreadsheet.

Goal maximized the value for the IT department instead of shareholders.

Lack of flexibility for choosing server upgrades.

No accounting of potential lost revenue for suboptimal datacenter performance.

Unrealistic projections of demand for datacenter services.

Provisdom's Solution

Required only 3 days for complete analysis.

Straightforward inclusion of uncertainty for future demand and revenue.

Flexible strategy based on projected learning about demand. The upgrade strategy balances expected future revenue with upgrade costs based on current information about demand.

Proper inclusion of tax and depreciation effects on risk-adjusted economic value.

Direct maximization of risk-adjusted economic value instead of departmentally motivated goal.

Results

Identified spreadsheet errors causing discrepancies of between $119 million and $172 million in risk-adjusted economic value.

Included more information known to be relevant, providing a seven-year strategy increasing risk-adjusted economic value by $460 million.

The strategy would also prevent an average of 66,000 metric tons of carbon dioxide emissions, while still meeting 100% of the data demand even in extreme cases.

Time required: 3 days + 3 hours of client time.