Guided Benchmark

This is a part of series of posts and the first two parts are “BI Essentials: Benchmarks” and “Benchmark Types“.

Second type of benchmark is universal in implementation, which nowadays has become a basic requirement in any project. Common terms to refer these benchmarks are Budget, Target and Forecast. At the outset, all three terms seem to have the same meaning, Interpretations do vary in subtle manner depending on nature of organization process, operating industry and functional area. Predominantly, the meanings are as follows,

  • Budget typically indicates the “Maximum” limit set for a measure
  • Target usually represents the “Minimum” limit set for a measure
  • Forecast value show a metric’s “Actual” value co-mingled with with a planned measure

The process of generating the guided benchmarks also vary greatly between organizations and departments. Planning process can be as simple as calculating the numbers in an spread sheet or utilize ready made planning tools in the market that incorporate complex algorithms.

In BI context, the guided benchmarks are treated as independent measures, among which Budget and Target are typically mutually exclusive. Benchmark_6For reporting needs, the benchmark measure has to be compared with an actual value in order to give user the position on where the organization stands. In some instances actual can be projected along with two guided benchmarks and beyond two, it becomes a clutter to interpret the data within the UI.

In all cases, a single dimensional element should be the baseline for guided benchmarks e.g. Profit Margin Current Vs Target for Product X or Travel Expenses Actual Vs Budget for Marketing Team. The second example illustrates one common feature of benchmarks encountered regularly, where the values are set for a group rather than individual dimension element.

In some instances, the benchmark will have multiple versions of the same. This is evident in “Forecast” type, where multiple versions are generated based on

  • Multiple scenarios, e.g. Forecast V1 indicates values when Market Growth is 5% and Forecast V2 indicates values when Market Growth is -3%
  • Merged actuals, e.g. at the end of first quarter, forecast values are overwritten with actual values from Jan to Mar and re-planned, which will be added as a separate version  in the system

Guided benchmarks are performance improvement measures and users interpret the values slightly differently from self benchmarks. Here, comparative value is known and the thought trail is typically

  • Are we in good or bad shape?
  • Should we implement any changes to day to day business activities?
  • What changes have to be implemented?

Note: Planning function, by nature is a complex topic with high degree of variability attached to it. Preparing and adopting guided benchmarks is a complex process by itself with several people across the organization involved. Planning teams collect historical data and perform analysis and bring all stakeholders into agreement before publishing the benchmarks. Entire processed gets revisited every year/quarter/month with guided benchmarks varying as per requirements (e.g. Target of 1000 widgets to be sold for Jan 2014 does not essentially mean that target for Jan 2015 will be the same and it can be any value between 0 to N).