What Makes Product Portfolio Planning More Efficient?

There’s no catastrophic event. One badly chosen product does not bring down a portfolio of products, but there is a creeping ramp up in commitments to launch, all of which have been analyzed in isolation, and so all engineering, tooling and launch marketing capacity has been fully subscribed and then it becomes hard to cut back any of the committed programs while none are delivering. Efficient portfolio analysis is not about the volume of analysis but rather about establishing a common basis on which to measure and compare committed programs.

Score against capacity, not against each other

First rank all programs by their expected value and risk. Then apply the “filter” of a number of constraints like a maximum of 6 industrialization launches supported or 4 new products to be validated through qualification programs per quarter with the lab capacity already in place. Twenty programs with high expected margins have little value if only a few of them will launch as industrialized products.

Assume for your modeling 2-3 key scarce resources in your organization. This may be far from headcount and so very different from the typical scoring against a ‘prioritize-then-cut’ style scorecard. In manufacturing, examples of such scarce resources are, the number of regulatory submittals that can be processed; the number of test rigs in service; the number of supplier qualification processes that can be processed.

Making the cut line visible

Once you have created a model of the load on the 2-3 key scarce resources in your organization, you draw a line where the capacity will run out. All projects below this line are unfunded until further capacity becomes available. A very practical way to discuss displacement rather than trying to determine whether or not a project is good enough to fund.

Standardize the inputs before you standardize the decisions

The largest number of errors in portfolio reviews result from too dissimilar input data (Volumes in Units or $, Tooling in Cost of Goods Sold or Capital).

Minimum set of input data, This must be the same for all business units and as complete as possible to enable a comparison of different proposals.

• Volume forecast by year with a stated confidence range, not a single number

• Full landed cost estimate including tooling, qualification, and launch inventory

• Named internal constraints the project will consume, with duration

• The decision date after which delay materially reduces value

• What the project replaces, extends, or cannibalizes in the existing range

A particular item is often missing from Portfolio Reviews, the deletion of a Portfolio Item. While new items are discussed one by one for addition to the portfolio, previously added items are never removed from consideration for the portfolio.

Give every function the same picture at the same time

Cross-team visibility is a communication problem and not a separate problem to solve. Sales plan against forecast updated monthly. Operations plan against schedule updated weekly. Finance plan against quarterly accounting consolidation updated on a quarterly basis using the same versions of the portfolio as the other functions.

By building a company’s distinct processes for managing all projects (under consideration as well as committed), such as cost planning for proposed products and services for which investment decisions have not yet been taken, into a single data repository, well-chosen portfolio planning software for manufacturing gives every function one place to monitor status across all the projects and products a business is considering.

Review cadence that matches decision type

This does not necessarily have to be the same group of people that made the initial decision to include a particular product in the portfolio. Organizing distinct decisions is easier than organizing a monthly review of the whole portfolio.

Connect the plan to the factory it depends on

Inefficient portfolio planning leads to products not being discontinued in time. This is because the cost of keeping a product in service is thinly spread across an organization and therefore invisible. The immediate cost of discontinuing a product is highly visible. Service parts, service documentation and engineering time to maintain products in service are all consumed by a long tail of products that are never accounted for in a portfolio plan.

These should be built into the plan as dated dependencies rather than assumptions. Three simple things to check before approving a plan.

• Does any quarter contain more line changeovers than the plants can absorb without losing output?

• Are two or more projects depending on the same supplier qualification slot?

• Does the capital spend profile match the phasing finance has actually approved, not the total?

Treat slippage as portfolio data

Once the released capacity is clear, it can be reallocated to other alternative Projects as soon as possible. The difference between actual and planned consumption of a resource over several cycles is also very valuable to improve the accuracy of future estimates much more than trying to get a better estimate in the first place.

Retire products with the same rigor you launch them

Sadly, the cost of maintaining a product through its service life is spread throughout the organization, whereas the cost of discontinuing a product is immediate and “embalmed” in the organization’s financials. Thus, long tails of poor or even obsolete products and related Service Parts, documentation and Engineering time are not even accounted for in a typical portfolio plan.

Set up a standing review threshold of, e.g., a contribution below a certain level for 4 quarters or more in succession and force the review of these ‘hopeless cases’ onto the agenda without having to request the review. Efficient portfolio management is having to cut as many inefficient products and measures as possible.