This Guest paper was submitted for publication and is copyright to Gary J. Summers © 2009
Published October 2009

Introduction | The Bayes' Law PPM Model | Proposals and Selection
New PPM Metrics | Management | Improving Your PPM Situation | Conclusion

New PPM Metrics

The Bayes' law model is general, but the analysis of your PPM results must match your method of selecting proposals. I am developing an analysis that compliments the most popular method of selecting proposals: funding down a ranking. If you use a variation of this method, as described above, and if my research is successful, your PPM results can produce the PPM metrics I describe below.

Evaluating Your Proposals and Prioritization

Figures 4 and 5 are metrics that estimate P-proposals and QPS. From this example one learns that:

  • The excellent prioritization for office inkjet printers is wasted because proposal processes produce few good ideas for this business line.
  • The company can invest heavily in its line of office laser printers. With good prioritization and proposal processes, aggressive selection will produce a high value for P-results, which will produce strong financial results.
  • Aggressive selection in the professional printing line will produce poor portfolios. If professional printing is strategically important, the company must invest to improve its proposal processes (P-proposals) and prioritization

Strategic Bucket

P Proposals

Office inkjet printers


Office laser printers


Professional printing


Figure 4: Table listing % of Good proposals in each strategic bucket
Figure 5: The quality of prioritization and project selection for each strategic bucket
Figure 5: The quality of prioritization and project selection for each strategic bucket

Cutoff Values, Hurdle Rates and Bucket Sizes

The preceding metrics, and forthcoming Figures 8 and Figure 9, arise from your proposal evaluations and a twofold classification of completed projects: Good vs. Bad. With a more thorough evaluation of completed projects the model can provide additional metrics. Specifically, if you can estimate the value created or lost by each project, the new model can relate project selection to portfolio ROI and NPV. Figures 6 and 7 illustrate these metrics. (For ease of presentation, these Figures assume that projects are evaluated on a ten-point scale and that Bad projects have a NPV < 0.)

Figure 6: How NPV varies with cutoff values
Figure 6: How NPV varies with cutoff values

For each strategic bucket, Figure 6 shows how NPV varies with the cutoff value that you select. If you select a high cutoff value, P-results will be high, so NPV > 0. However, you will select few proposals, so NPV will be small. If you lower the cutoff value and select more projects, P-results will decrease. Initially, the value from the additional Good projects will exceed the losses from the additional Bad projects, so NPV will increase. However, if you continue to lower the cutoff value, will decrease too much. At some point, the losses from the additional Bad projects will exceed the value created by the additional good ones. When this situation occurs, NPV will decrease. If you continue to reduce the cutoff value, NPV will become negative.

The maximum of each curve shows the optimal cutoff value for its strategic bucket. The points where NVP = 0 show the lowest cutoff values you can set before each bucket's NPV turns negative. Figure 6 could display NPV vs. hurdle rate or bucket size (budget), so the new model can help you set cutoff values, hurdle rates and bucket sizes.

Figure 6 is analogous to the productivity curves found in PPM dashboards and the "efficient frontier" produced by optimization. However, Figure 6 differs from those metrics in two ways. First, Figure 6 considers the impact of prioritization errors, but the other curves do not. For this reason, productivity curves and "efficient frontiers" display an optimistic bias. Second, productivity curves and "efficient frontiers" present expectations that arise from the assessments of your current proposals. Figure 6 presents expectations that arise from your past portfolios. It reveals your company's proven performance. Both types of metrics are useful and should be considered.

Balancing Strategic and Financial Goals

Figure 7 reveals the relationship between each strategic bucket's cutoff value and ROI. A high cutoff value (cautious selection) raises ROI. A low cutoff value (aggressive selection) lowers ROI. This relationship occurs for two reasons. First, in a prioritized list the "bang-for-the-buck" tends to decrease as one moves down the ranking. Second, because prioritization is imperfect, P-results decreases as one selects down a ranking. This relationship implies a trade-off between strategic and financial goals. In the short-run, aggressive pursuit of strategic goals will hurt financial performance.

Figure 7: How ROI varies with cutoff values
Figure 7: How ROI varies with cutoff values

Figure 7 helps you manage this trade-off. To see how, suppose you wish to obtain 30% of revenue from new products while producing an ROI = 15%. You can estimate the number of proposals you must select to produce the revenues. This estimate identifies the cutoff value that will achieve your strategic goal. With Figure 7 you then estimate the ROI for this cutoff value. If the estimated ROI < 15%, your strategic and financial goals are incompatible and destined to fail. If this situation occurs, Figure 7 will help you select the best trade-off for your company. You can then relax the trade-off by using methods I present later in this paper.

Proposals and Selection  Proposals and Selection

Home | Issacons | PM Glossary | Papers & Books | Max's Musings
Guest Articles | Contact Info | Search My Site | Site Map | Top of Page