TVP – Metric 41 Probability of Success

Background | User Guide | Program Contents | Stakeholders | List of Metrics

1. Metric Definition

Probability of Success characterizes the breadth and depth of the new product pipeline. It can also be used as a key criterion for passage through gates in the stage-and-gate process.

The probability of success for an R&D program, or project, depends on the likelihood that technical problems can be resolved (probability of technical success) and that the market opportunity can be realized (probability of commercial success). The product of these two factors can be used as an estimate of the overall probability of success of the project.

Typically, a business or company will want to have a portfolio that is balanced in terms of high, medium and low risk projects in a manner which is consistent with corporate strategy of that company.  This balance assumes the most likely scenario, namely, that risk is inversely associated with the potential returns of the project.  The probability of success of a program will also be used as a key criterion for passage through the gates of a stage-and-gate management process. The probability should remain constant or increase as development proceeds and as successive gates are reached. Projects whose probability of success decreases over time should be closely evaluated and most likely discontinued.

2. Advantages and Limitations

Once an effective system is established, this metric is among the best predictors of R&D’s future success, and is invaluable in analysis of portfolio balance.  Measured over time, it provides a good measure of the organization’s predictive ability.  It can also serve as a warning alarm for problems and fears not yet articulated.

A disadvantage of the metric that it is qualitative and subjective.  It is also subject to “gamesmanship” or misuse, possibly being used to push favorite projects through the system. The metric requires solid market knowledge and forecasting ability, with early market forecasts sometime becoming overly relied upon, or “engraved in stone.”  It is also subject to technical overconfidence.  Early stage predictions can become difficult to change through the lifetime of the project due to a sense of disagreement or conflict.

3. How to use this metric

An informal panel of program/project members and other internal and external experts should review the program, potential technical issues and obstacles impeding its progress. Each panel member should rate the likelihood of project success (including meeting all product attributes) on a scale of one to ten, where ten is “very simple to complete” and one represents  “insurmountable obstacles present”.  Marketing personnel should regularly estimate the probability of commercial success using the same numerical system. The product of the two factors is the overall success probability.  Significant changes, particularly drops in attractiveness, should be investigated for underlying causes. The cumulative probability-adjusted projected peak sales generated by the R&D portfolio is a gauge of the organization’s new product prospects.

4. Options

The projected, probability-corrected sales forecast spread over five to ten years is a valuable measure of the R&D new product pipeline. Gaps are easy to spot and remedial adjustments in the portfolio can be made. Probability trends may be used as an indication of enthusiasm or lack of enthusiasm for a specific R&D program.  Calculations can be refined using net present value calculations, but the rough nature of the estimates at early stages usually negates the effort. 

(This option is essentially the same as the metric “Projected Value of the R&D Pipeline”)

5. Champions and Contacts

6. References

Merrifield, D. Bruce, 1981, “Selecting Projects for Commercial Success”, Research Management, 24 (6), 13-18. This is an early source article for the technical probability of commercial success.

Ellis, L. W. 1984. The Financial Side of Industrial Research Management. New York: Wiley. Chapter 6 discusses the probability distribution of outcomes.