Software Development Governance - POT: Making the Best Investment Choices Across a Portfolio

Funding managers must determine how to allocate available funds to different products in their portfolio to have the best chance of achieving their financial objectives, without exposing the organization to unacceptable levels of financial risk.

The Portfolio Optimization Tool (POT) automatically identifies funding allocations that offer the best chance of achieving an organization’s financial goals, within the organization’s ability to tolerate risk.

  • POT works on financial models created with Financier. Funding managers need only supply their budget and financial objectives.
  • What-if analysis and alternative comparison features enable funding managers to evaluate the relative cost, benefit, and risk of different funding allocations before making a decision.

An Illustrative Example

Suppose Mike is in charge of the software project portfolio in a small size software firm. At the beginning of the year, he is going to decide how to allocate this year's investment to the projects proposed in the portfolio. The following chart illustrates the usage scenario:


Portfolio Modeling

Mike inputs the portfolio data into POT via its dedicated GUI, including the porfolio name, description and investment start and end times. He also enters information about his investment proposal: proposal name, description, and proposed feature sets.

Cost/Benefit Estimation

Next, Mike enters the estimated benefit and cost stream for each feature set. For each proposed feature set, three points on a time grid denote the low value, likely value and high value of the expected benefit or cost:

Optimization Condition Setting

Mike now selects one of several predefined optimization patterns (e.g., maximize the worst case NPV) and sets the appropriate parameters. As shown below, the selected pattern is illustrated with a diagrammatic presentation so that a Mike can easily understand the meaning of the relevant parameters for the pattern. For example, Mike sets the budget number to $900,000 and requires that the expected value of the "5% higher tail region" should not exceed the budget.

Generate and Review Optimal Solutions

Mike launches the automatic optimization by clicking "Optimize Funding" and POT returns an optimal solution as shown below. Mike reviews the constituents in the optimal solution, the solution's NPV, total benefit, total cost, and ROI.

Mike may change the optimization condition or switch to another optimization pattern to generate alternative optimazation solutions for comparison. Alternate solutions are shown in adjacent columns.

What-if analysis

Mike might have a subjective solution in mind and launches the what-if analysis function to explore and compare. In what-if mode he can make adjustment to a previously generated solution, and add or remove feature sets to generate a new solution. He can then review the NPV/total benefit/total cost gap between solutions. The manually generated solution can be saved for further comparison.

Visual comparison

Mike opens the solution pool and selects a set of solutions for comparison. There are several charts he can use for different purposes. The "Expected Value" chart provides comparison based on mean NPV, Benefit, or Cost. The "Distribution" chart provides comparison based on the distribution of NPV, Benefit, or Cost. The "NPV vs. Risk" chart provides comparison based on the mean of NPV and risk.

POT is a Research prototype. It works in conjunction with Financier to provide comprehensive support for managing investments in a portfolio.