The BCG growth-share matrix and portfolio decisions

Any business with more than one product, brand or unit has to decide where to put its money. Some parts need cash to grow. Some produce more cash than they can use. Some should probably be sold or closed. The BCG growth-share matrix is a simple picture that helps with those decisions, and it has been used for more than fifty years.

This guide explains how the matrix works, how to build one, what it says about portfolio choices, and the limits BCG itself has pointed out.

Where the matrix comes from

According to BCG’s own history, the matrix was a joint effort inside the firm. Alan Zakon first sketched it, colleagues refined it, and the founder, Bruce Henderson, made it widely known in his 1970 essay The Product Portfolio. (The same BCG page gives 1968 in its metadata, so sources differ on the exact year the idea was first drawn; 1970 is the date of the published essay.)

Henderson’s argument was about cash. He wrote that “high margins and high market share go together”, which he linked to the experience curve: the firm that produces most tends to have the lowest costs. Growth uses cash, because it needs investment in assets, and holding share in a fast-growing market needs more. He also noted that “no product market can grow indefinitely”. The return from growth comes when growth slows.

How the matrix works

The matrix plots each business on two measures:

• Market growth rate, a measure of how attractive the market is and how much cash the business will need.
• Relative market share, usually your share divided by the share of the largest competitor, a measure of competitive strength and cash generation.

BCG’s history page sums up the recommended actions: milk cash cows for cash to reinvest, invest significantly in stars, invest in or discard question marks depending on their chance of becoming stars, and liquidate, divest or reposition pets. Henderson was blunt about pets. He wrote that they are “essentially worthless, except in liquidation”, and evidence of failing either to win leadership or to get out in time.

Building a matrix step by step

1. Choose the units. Each point should be a business, product line or brand with its own market and competitors.
2. Define each market carefully. Share depends entirely on the market definition. A brand can be a leader in “vegan protein bars in the UK” and tiny in “snacks”.
3. Estimate market growth for each unit, ideally the expected growth over the next few years, not only last year’s figure.
4. Calculate relative share. Divide your share by the largest rival’s. A value above 1 means you lead.
5. Set the dividing lines. State the growth rate and relative share you use to split high from low, and say why.
6. Plot and size the points. Make each bubble proportional to the unit’s revenue so the picture shows where the money is.
7. Add cash flow. Next to each bubble, note whether it generates or uses cash today.

Turning the picture into portfolio decisions

Henderson’s conclusion was that a company needs a balanced portfolio: cash cows to fund the business, stars to become future cash cows, and question marks to become future stars. The matrix helps you ask:

• Do our cash cows produce enough cash to fund our stars and the question marks we want to back?
• Which question marks have a realistic path to leadership, and which should we stop funding?
• Are we keeping pets out of habit or sentiment?
• When today’s cash cows decline, what will replace them?

Treat the answers as hypotheses to check, not as verdicts. A unit in the pet box may still have a role, for example if customers expect it as part of a range.

A worked illustration

An example (illustration only). A small UK food company has four product lines.

The picture suggests using chutney cash to keep chilli oils ahead while that market grows. For plant-based dips, the team sets a test: a defined budget and twelve months to reach a clear share target, or they exit. Gift hampers take up warehouse space and management time out of proportion to their sales, so the team plans to wind them down after the next Christmas season. Each of these is a decision to review, not one the chart makes alone.

The limits, including BCG’s own update

In 2014 BCG published a review of its own matrix. It found that business conditions had changed in ways that weaken the original assumptions:

• In 75 percent of industries, the average time a business spent in one quadrant fell from four years in 1992 to under two years in 2012.
• Cash cows captured 25 percent less of total profits in 2012 than in 1982.
• The authors concluded that “market share is no longer a direct predictor of sustained performance”.

BCG did not drop the matrix. It recommended reviewing the portfolio more often, running more question marks as cheap, fast experiments, keeping pets “on a short leash”, and selecting winners more carefully with better data. It also suggested that market share may need replacing with a new measure of competitiveness on the horizontal axis.

Other limits are practical. The result depends on how you define each market. Growth is only one sign of attractiveness; a slow market can still be very profitable. The matrix also ignores links between units, such as a pet that brings customers to a star.

Common mistakes and when to use it

• Using absolute share instead of relative share. A 20 percent share means leadership in one market and a weak position in another.
• Drawing the matrix once. BCG’s own data says positions change faster than they used to.
• Applying it to a single-product business. The matrix is about a portfolio. With one product, an industry analysis is more useful.
• Letting the labels decide. “Dog” sounds final, but the decision should rest on cash flow, strategy and links to other units.

Use the matrix for a first, visual conversation about where cash comes from and where it goes, for annual planning across several units, or to challenge a portfolio that has grown by accident. Pair it with a proper look at each market’s profitability and with a review every few months rather than every few years.