Portfolio analysis is a method for systematically rating all the projects, business units or investments in an organisation — so you can decide where money, time and people are best spent.
It makes visible what creates value and what merely ties up resources, giving you the basis for investment and prioritisation decisions.
In this article: what portfolio analysis is, the important difference between a financial and a project portfolio, the best-known models compared, a step-by-step guide and a concrete example from agency life.
What is portfolio analysis?
Portfolio analysis answers one hard question: what makes us money — and what doesn't?
Instead of looking at each project or investment on its own, you look at the whole bundle (the "portfolio") and rate every part against the same criteria — typically by return and risk, or by market attractiveness and your own strength. That way you can see at a glance where to invest, hold or exit.
Portfolio analysis is used by management teams, asset and fund managers, private investors — and in project management by everyone who runs several projects in parallel and has to spread resources sensibly.
Financial portfolio vs project portfolio: the difference that's often missing
The term "portfolio analysis" gets used for two related but different things. Telling them apart is what makes you apply the models correctly:
- Financial portfolio analysis: rating investments and assets (shares, funds, holdings) by return and risk. The goal is to optimise an investment portfolio.
- Project portfolio analysis: rating all of an organisation's current and planned projects. The goal is to do the right projects — and drop the wrong ones. This isn't about share prices, but about margin, utilisation, strategic fit and risk per project or client.
For teams that deliver projects, it's usually the second one that's meant. The classic models come from the strategy and finance world, but they map straight onto a project portfolio.
The most important methods and models
There isn't one portfolio analysis. Three models shape the practice — each with its own focus.
Model: BCG matrix
Rates by: Market share × market growth
Strength: Fast, intuitive, four clear categories
Typical use: Sorting products and business units roughly
Model: McKinsey matrix
Rates by: Market attractiveness × competitive strength
Strength: More nuanced, nine cells, more factors
Typical use: More detailed investment decisions
Model: Markowitz model
Rates by: Return × risk
Strength: Mathematically grounded, diversification
Typical use: Financial and investment portfolios
BCG matrix (growth–share matrix)
The best-known model, developed by the Boston Consulting Group. It sorts each unit into one of four cells based on relative market share and market growth:
- Stars — high market share, high growth. Invest here.
- Cash Cows — high market share, low growth. Milk what you can.
- Question Marks — low market share, high growth. Potential unclear — watch them or back them deliberately.
- Dogs — low market share, low growth. Candidates for an exit.
The charm of the BCG matrix is its simplicity. That's also its weakness: two axes rarely tell the whole truth.
McKinsey matrix (nine-box matrix)
The extended version. It rates each unit on two axes — market attractiveness (growth, potential, competition, profitability) and competitive strength (market share, quality, brand, innovation) — and places it in a grid of nine cells.
The result is more nuanced than the BCG matrix and suits decisions where more than two factors count.
Markowitz model (modern portfolio theory)
The mathematical model from Harry Markowitz — the foundation of modern investing. The core idea: investors are risk-averse and want the maximum return for a given level of risk.
Through diversification (spreading across different assets) you can lower risk without giving up return. Strong for financial portfolios, less so for projects.
How to do a portfolio analysis
In five steps — whether for business units, investments or your project portfolio:
- Define the portfolio. Decide what you're rating: all active projects, all clients, all product lines. One clear boundary, no mixing.
- Set the criteria. Pick two to four rating axes that fit your goal — for example contribution margin, strategic fit, risk, growth potential.
- Gather the data. Get real numbers instead of gut feeling: revenue, costs, utilisation, margin. This phase decides the quality of the whole analysis.
- Rate and map. Place each element in the matrix. Now it's visible what's a Star, a Cash Cow or a Dog.
- Decide and act. Derive concrete moves: grow, hold, rebuild, end. An analysis without a decision is just a pretty chart.
Example: portfolio analysis in an agency
An agency of 40 people runs 18 active client projects. It feels like everything's just "somehow" working. Only a portfolio analysis along the axes of contribution margin and growth potential reveals the reality:
- 3 Stars — large, growing clients with margins above 30%. Worth putting your best people on.
- 5 Cash Cows — stable retainers with reliable margins but little growth. Keep them efficient, don't over-tend them.
- 6 Question Marks — new clients with potential but thin margins. Look closely: develop or let go?
- 4 Dogs — projects that eat a lot of capacity and barely deliver any margin. Adjust prices or end them.
The penny drops: those four Dogs tie up almost a fifth of the team's capacity — time the Stars don't get. Without the analysis, no one would ever have spotted it.
Benefits and limits of portfolio analysis
What it gives you:
- A clear basis for decisions on investment, prioritisation and resourcing — driven by data, not gut feeling.
- Ongoing steering: rate your portfolio regularly and you spot trends and trouble early enough to act.
- Risk management through diversification and an honest look at what isn't working.
What to keep in mind:
- It simplifies reality. Two or three axes never capture every relevant factor. The matrix is a tool for thinking, not a machine for deciding.
- It leans on hard numbers. Qualitative aspects — team fit, strategic importance, the client relationship — easily fall through the cracks.
- It's not a crystal ball. Political or economic shocks and other external events can't be calculated away.
Why good data decides everything
A portfolio analysis is only as good as the numbers it stands on. And this is exactly where most fall down: margin, utilisation and project status live in separate tools, in spreadsheets, in people's heads. What sits in a spreadsheet as a snapshot is already out of date tomorrow — and an analysis built on old data leads to expensive wrong calls.
Teams rarely lose money through bad work, but through a lack of visibility. Only when you can see capacity, billability and project status in real time can you rate your portfolio at all.
In awork, projects, time tracking and utilisation come together in one system — the data foundation that turns a portfolio analysis from a nice diagram into a real steering tool.
Frequently asked questions about portfolio analysis
What is portfolio analysis in simple terms?
A method for rating all of an organisation's projects, units or investments against the same criteria — usually return and risk — so you can decide where it's worth putting your resources.
Which portfolio analysis models are there?
The three best-known are the BCG matrix (market share × growth), the McKinsey matrix (market attractiveness × competitive strength) and the Markowitz model (return × risk) for financial portfolios.
What is the BCG matrix used for?
To quickly sort business units or products into four categories — Stars, Cash Cows, Question Marks and Dogs — and derive investment decisions from them.
How do I do a portfolio analysis?
In five steps: define the portfolio, set the rating criteria, gather real data, map each element in the matrix, then derive and carry out the actions.
What's the difference between financial and project portfolio analysis?
The financial version rates investments by return and risk. Project portfolio analysis rates projects by margin, utilisation, strategic fit and risk, so you can prioritise the right ones.
