What to remember
- Define units and the decision before scoring.
- Choose the matrix for the question it supports.
- Keep priority, feasibility and allocation separate.
- Record trade-offs and review conditions.
Define the portfolio and expected decision
A portfolio may contain businesses, offers, markets or initiatives, but units must share an analytical level. Identify duplicates, dependencies and mutually exclusive options before scoring.
Specify the output: invest, maintain, test, defer, divest or stop, with horizon and resource envelope. Without that decision, a ranking is only an ordered list.
Choose the framework for the question
Ansoff classifies a growth direction. BCG reads growth and relative share. GE-McKinsey builds a multi-factor view of attractiveness and strength. A multi-criteria matrix compares options across defined objectives.
Frameworks may follow one another, but their scores should not be mechanically merged. Keep each method’s units, assumptions and limits.
| Question | Primary framework | Output |
|---|---|---|
| What type of growth? | Ansoff | Product-market category |
| What portfolio position? | BCG | Growth-share diagnosis |
| What multi-factor priority? | GE-McKinsey | Attractiveness-strength zone |
| What trade-off between options? | Multi-criteria | Comparison and sensitivity |
Test priorities against constraints and interactions
Priority is not allocation. Budget, people, schedule, eligibility thresholds, risk, dependencies and synergies determine feasible combinations.
Portfolio-selection research separates evaluating items from composing a feasible set. A high-ranked project may wait for missing capacity; two medium options may create joint value that separate scores miss.
- Budget and cash
- Skills and capacity
- Dependencies and exclusions
- Synergies
- Risk and reversibility
- Sequence and milestones
Record allocation and review
The target workflow links the decision record to Atlas, portfolio structure to Grid, analysis and synthesis to Plot, and downstream price-volume-cost quantification to Mix. It describes editorial coherence, not a working public end-to-end chain.
For each unit, record the decision, amount or capacity, reasons, trade-offs, next evidence and event that reopens the decision. Accountability remains human.
Frequently asked questions
Must the whole portfolio use one score?
No. A common core helps, but hard constraints, interactions and maturity differences may remain outside the score.
Can BCG and GE-McKinsey be combined?
Yes as successive readings, provided their scales and assumptions remain separate.
When should allocation be reviewed?
At a useful date and when a predefined signal invalidates an assumption, capacity or constraint.
Full sources
- Strategy-oriented portfolio selection and dynamic synergiesComplete source available online · Read the complete source
- Priority-based portfolio selectionComplete source available online · Read the complete source
- Portfolio selection under risk and uncertain incomeComplete source available online · Read the complete source
- Prioritizing projects in a portfolioComplete source available online · Read the complete source