Almost every manufacturing business has more great ideas than it has the people, test time, tooling dollars or supplier attention to transform those ideas into physical products to ship. As a result, prioritization isn’t about picking winners, it’s about making the best decisions with respect to how to allocate limited resources across competing demands.
Begin with the binding constraint, not the candidate list
Ranking projects before understanding their limitations as to throughput can result in plans that look great on paper but fail to meet promised delivery dates. More importantly, determining which of your limited resources will run out first when you approve all of your ideas is critical. Typically, this will be validation and test capacity, or possibly the small group of legacy engineers that understand the current process, or the capital tied up in existing tooling with 12 month lead times.
Name the capacity that actually limits output
Rather than think about the value of individual projects, measure how much of the constrained resource each will consume and what return you will get on that. Track the projects in terms of qualified test weeks, additional mold cavities, new regulatory submission slots per year etc.
Choose a scoring method that matches your data
The scoring method fails when the numbers are expected to be more precise than the underlying data allows. A 5 year NPV for a derivative in a stable category is a reasonable number to consider. A 5 year NPV for a new category is pure theater until an installed base develops.
| Method | Best suited to | Inputs required | Common failure |
| Weighted scoring | Mixed portfolios with many small projects | Agreed criteria and weights | Scores cluster, nothing is deprioritized |
| Return per constrained unit | Capacity limited engineering or test groups | Margin forecast, capacity consumption | Ignores strategic or compliance work |
| Cost of delay | Time sensitive launches and tenders | Revenue timing, competitor and contract dates | Overstated urgency on every item |
| Strategic buckets | Balancing core, adjacent and new bets | Executive agreement on allocation | Buckets defended long after the strategy shifts |
Separate the categories that should not compete
One has to be careful not to confuse required production runs with future growth opportunities. The former have to be scheduled as committed work and their capacity needs to be subtracted from the total available. The resulting portfolio can then be ranked for the available capacity. Mixing mandatory production runs with future growth initiatives leads to the former always winning out over the latter in ranking.
Give every function the same view of the pipeline
One of the bigger problems of ranking projects to invest in within a manufacturing organization is that typically different groups have their own views of reality. Therefore salespeople might track opportunities and marketing initiatives, the engineering organization might track the projects that they need to bring to market and the operations people might track all the part numbers. Until there is a launch delay, these different views of reality are never reconciled. If you want those views to converge, it is worth adopting software built for product portfolio planning for manufacturers so that sales, engineering and operations are reading the same pipeline.
Agree what a stage actually means
Ensure entry and exit criteria for each stage are written down and consistently applied. A project at design validation should have fixed requirements and a supplier committed to delivering said requirements against said fixed requirements, not ‘we are waiting to hear back from supplier X’. Seeing stage labels actually mean something on the aggregate view of your portfolio allows you to use the ranked list as a proper planning tool.
- One owner per project, accountable across functions rather than within one
- Capacity shown as allocated versus available, not as a count of active projects
- Assumptions recorded with the decision, so later reviews can test them
- A single refresh cadence, with changes visible to everyone on the same day
Sequence against manufacturing lead times
Even the best ranked list is only a plan if the sequence of launches through the manufacturing environment takes into account key factors such as tooling lead times, supplier qualification time, line changeover time and stockout of existing products. Sequencing is critical to launch date and to profitability in the first year of launch.
Work backward from the slowest commitment
Identify the longest lead component decision in each project and schedule the approval gate for that component first. This ensures that you are approving to start to test, to begin production, to phase in and to phase out in a manner that is consistent with your overall production goals and corresponds to your start-up plans.
Build stopping rules before you need them
A portfolio of approved projects also drifts, because starting projects is easier than stopping them once they have been approved. Therefore, it is also necessary to define the stop criteria for a project already approved, and to review these same criteria at each subsequent gate as one reviews the project’s progress.
- Set a tripwire for each major assumption, such as landed cost exceeding a stated threshold.
- Require a named destination for released capacity, so stopping funds something else.
- Rebalance quarterly against the constraint, not annually against the budget cycle.
- Track forecast accuracy by project type, then adjust the weightings accordingly.
What matters most is how you apply any method to assess potential projects in terms of tested assumptions, the required time and the available capacity. Sticking to that is more important than the chosen method. Keep updating it.
