Making Better Use of Metrics

Have more questions? Submit a request.

Many Decision Lens customers express a desire to become much more data-driven in their decision-making, but often lack a full set of reliable quantitative performance data to support this. So decision teams often fall back on qualitative estimates of benefit to complete their decision. The good news is that Decision Lens can accommodate both subjective and objective (quantitative) criteria in prioritization and optimization decisions.

However, because stakeholders generally are very familiar and comfortable with performance metrics in their respective fields, it is a very good analytics strategy to incorporate performance metrics into decision models to help communicate the scale and impact of portfolio decisions in key areas. Decision Lens offers simple ways to leverage quantitative data in decision models through the use of metrics.

Alternative Metrics

An alternative metric in Decision Lens is a quantitative attribute associated with an alternative (and displayed as a column in the Alternatives page). Examples of this include:

  • The NPV, ROI, and Payback Period of an IT project that has revenue and cost impact
  • The forecast reduction in crash rates for a section of highway if a specific rebuilding project is implemented
  • The operating cost savings associated with adopting a new formulation for a food product 

Each of these examples describes a specific measurable benefit associated with a project one might find in an IT, transportation, or food & beverage R&D portfolio (respectively).

Any criterion that is assigned a linear numeric scale automatically creates a corresponding metric of the same name in the Alternatives screen. While the criterion rating captures the rating value and converts it to a 0-1 score on the scale, the metric captures the input value. For example, an IT project with a forecast NPV of $5M might receive a criterion score of .25 (assume the highest NPV project in the portfolio is $20M), while its metric value is 5 (or 5,000,000).

The use of metrics is not limited to informing quantitative criteria during ratings. Here are some ways to incorporate more "hard" data into decisions:

  • "Tie Breaker" Measures: In many cases, simple prioritizations of larger portfolios can result in "clumps" of alternatives with similar Priority Value scores. Adding simple performance metrics, for example project cost, ROI, lead time, etc., and displaying them as a column in Sensitivity Analysis can help stakeholders identify meaningful differences among alternatives peers where choices are not obvious.
  • Alternative Cost: Alternative cost data can be incorporated into decisions outside of Resource Balancer using metrics. This can help stakeholders in simple prioritization decisions consider cost without going through resource allocation. Simply create a metric for alternative cost and display it in Sensitivity Analysis or Bubble Chart visualizations.
  • Impact Dashboards: Impact Dashboards are a feature of Decision Lens that enable optimization across single or multiple portfolios using performance measures. For example, an IT organization may be seeking the optimal balance between projects that generate new digital business, and other projects that reduce legacy systems and the costs and risks they incur. These portfolios will be evaluated against differing criteria and metrics, such new revenue and NPV for digital business, and end users and maintenance backlog eliminated for legacy systems. Using Impact Dashboards, these four metrics would be used to optimize IT investment based on reaching specific targets in each area.

Metrics can be created, populated, tracked and reported as alternative attributes in a portfolio. They can also be reported as sums or averages in portfolios that have been optimized. And with Impact Dashboards, portfolio metrics can be used to optimize portfolios (vs. resource allocation).

Importing Metric Data

If you need to populate a large alternative list with metric data that is not associated with a criterion, there is an easy trick you can do.

  1. Create a temporary criterion at the top of your hierarchy.
  2. Assign the temporary criterion a linear quantitative rating scale. This will create a corresponding metric of the same name in the Alternatives list.
  3. Import the data from a spreadsheet as ratings data for the criterion.
  4. Delete the temporary criterion from your hierarchy. This will not impact the metric that was created, nor the data that was imported. It will stay in the Alternative list.


Was this article helpful?
0 out of 0 found this helpful


Powered by Zendesk