What a BCG matrix is
A BCG matrix, or growth-share matrix, is a two-by-two grid that plots each product line or business unit by how fast its market grows and how much of that market it holds against the leader. Its four quadrants, stars, cash cows, question marks and dogs, suggest where to invest, where to take cash and what to drop.
The Boston Consulting Group built it as a portfolio tool. It answers one question: with limited cash, which products should get more of it.
How the growth-share matrix works
Market growth, the vertical axis
The yearly growth of the market the product sells into, not of the product itself. Where to draw the line is a judgement. Many teams use about 10%, others use the growth of the wider category.
Relative market share, the horizontal axis
Your share divided by the share of your largest competitor. If you hold 12% and the leader holds 24%, your relative share is 0.5. If you are the leader, divide by the number two: 30% against 15% is 2.0. The line sits at 1.0. By convention high share is on the left, often on a log scale.
Relative share is the point, because the matrix assumes the leader makes the most cash per sale. 20% of a market means one thing when the next brand holds 5% and another when it holds 40%.
Stars
High growth, high share. You lead a market that is still growing. Stars earn a lot and spend a lot, because keeping the lead in a growing market costs stock, marketing and people. The usual advice is to invest to hold the lead.
Cash cows
Low growth, high share. You lead a market that has stopped growing. Cows earn more cash than they need. The usual advice is to protect them and spend their cash elsewhere.
Question marks
High growth, low share. The market is moving, and you are not leading it. Question marks eat cash and may never pay it back. The usual advice is to back a few hard and drop the rest.
Dogs
Low growth, low share. Keep them if they make money or do a job for the rest of the range. Otherwise sell or stop.
The intended flow: cash from the cows funds the chosen question marks, which become stars, which become cows when their market slows.
Three BCG matrix examples
All three companies below are hypothetical. The growth rates and shares are invented to show how placement works. None of them is measured.
A consumer electronics accessories brand
Four product lines. Phone cases sell into a market growing about 3% a year, and the brand holds 1.6 times the share of the next largest seller: a cash cow. Wireless chargers grow at 14% and the brand holds 1.3 times the next seller: a star. Earbuds grow at 12%, but the brand holds a fifth of the leader's share: a question mark. Cables grow at 1% with 0.4 times the leader: a dog.
Cases pay for everything. Chargers need stock and ad budget to keep their lead. The hard call is earbuds, a fast market against brands with far bigger budgets. The matrix does not say "invest in earbuds". It says earbuds will cost cash either way, so decide now whether to fund them to lead or stop. Cables stay if they cost nothing to carry.
A supplement brand's range
Whey protein sells into a market growing 4% a year with 1.2 times the next brand's share: a cash cow. A magnesium sleep powder grows at 11% with 1.4 times: a star. Creatine gummies grow at 22% with 0.3 times the leader: a question mark. A daily multivitamin grows at 2% with a quarter of the leader's share: a dog.
The multivitamin is the product the founders launched with. It sells steadily in a flat market where the leader holds four times its share. The label is not a verdict. If the multivitamin brings in subscribers who later buy the protein, it is doing a job the matrix cannot see.
A B2B software company's products
Core invoicing grows at 5% with twice the share of the next vendor: a cash cow. An AI bookkeeping assistant grows at 25% with 1.1 times: a star, just. A payments add-on grows at 18% with half the leader's share: a question mark. An expense app grows at 3% with 0.3 times: a dog.
Software shows the weak spot. These products are sold together, and the expense app may be why customers stay on invoicing. Market lines blur too. Is the assistant in bookkeeping software or in AI tools? The answer moves it from star to question mark.
| Product line | Market growth | Relative share | Quadrant |
|---|---|---|---|
| Phone cases | 3% | 1.6 | Cash cow |
| Wireless chargers | 14% | 1.3 | Star |
| Earbuds | 12% | 0.2 | Question mark |
| Cables | 1% | 0.4 | Dog |
| Whey protein | 4% | 1.2 | Cash cow |
| Magnesium sleep powder | 11% | 1.4 | Star |
| Creatine gummies | 22% | 0.3 | Question mark |
| Multivitamin | 2% | 0.25 | Dog |
| Invoicing | 5% | 2.0 | Cash cow |
| AI bookkeeping assistant | 25% | 1.1 | Star |
| Payments add-on | 18% | 0.5 | Question mark |
| Expense app | 3% | 0.3 | Dog |
What you can see of a competitor's portfolio
You cannot compute a competitor's relative market share from outside. You need their sales and the category's, and for most private brands neither is public. Kvasano does not measure market share, and nothing below is a substitute for it.
What public data can show is where a competitor puts its effort. That is the decision the matrix exists to inform.
- Which products they push in ads. A catalogue ad that rotates the whole store says little. A product with its own creative, running for months, is one they are funding. On Loop Earplugs, 29 of 50 ads analysed in detail were catalogue ads rotating store products with no discount (ads last collected 6 September 2026).
- Which products they discount. A cut on one product alone is being cleared or pushed. A cut spread across the range is a pricing decision. Loop changed 38 prices between 15 and 27 September 2026, all of them down, and 18 of them dropped by exactly $1.00, one change landing on 18 products rather than 18 verdicts.
- Which products they stop stocking. A product that sells out and stays out, or leaves the catalogue, is often a dog being retired. One reading proves nothing. Watch it across weeks.
Together these show what a competitor treats as a star or a dog, not what the market makes of it. The ad and price readings sit on Loop's benchmark. The notes on seeing every ad a competitor runs and on reading their price ladder cover each signal.
How to make one
- List the units that each have their own market: product lines or business units, not single products.
- Define each market narrowly enough that "the leader" means something.
- Find the growth of each market over two or three years, not one.
- Estimate your share and the largest competitor's, and divide.
- Plot them. Draw the share line at 1.0 and a growth line you can defend, and write down why.
- Size each dot by the revenue it brings you, so a big cow and a tiny dog do not look equal.
- Write one decision per unit, invest, hold, harvest or drop, and where its cash goes.
- Redraw it every year.
What it cannot tell you
It assumes share means profit. The idea is that the leader has the lowest cost per unit. Plenty of small brands earn well in a niche, and plenty of leaders buy their share with margin.
Growth is not the same as attractive. A fast market pulls in rivals, and the price war can arrive before the profit does.
It treats products as separate. Bundles, entry products and add-ons that keep customers are invisible, so a dog that feeds a cow gets marked for cutting.
The market definition decides the answer. Draw the market narrowly and you lead it. Draw it widely and you are a dog. Both can be defended.
It was built for groups choosing between businesses. For a single brand in one category, the four lines share one customer and one budget, and the quadrants are closer together than the grid suggests. It also says nothing about direction. For that, the Ansoff matrix sorts growth moves by risk.
FAQ
What are the four quadrants of the BCG matrix?
Stars (high growth, high share), cash cows (low growth, high share), question marks (high growth, low share) and dogs (low growth, low share). Question marks are sometimes called problem children.
How do you calculate relative market share?
Divide your market share by the share of your largest competitor. If you are the leader, divide by the second largest. Above 1.0 you lead, below 1.0 you do not.
What is the difference between the BCG matrix and the Ansoff matrix?
The BCG matrix sorts what you already sell, to decide where cash goes. The Ansoff matrix sorts where you could grow next, new products or new markets, by how risky each move is.
Is the BCG matrix still useful?
As a first sort and a shared language, yes. As a decision on its own, no. Check each quadrant against profit and against how the products depend on each other before you cut anything.
Pick one competitor. See what we find.
Your first benchmark is free. One domain, measured against you, no card and no call.