---
title: "BCG Matrix: Unilever's Ice Cream Split and Capital"
description: "The BCG matrix helps allocate capital between business units; it does not say which to drop. Unilever's ice-cream split, a four-unit example, four traps."
url: "https://staging.khoi.co.uk/en/blog/unilever-tach-kem-ma-tran-bcg"
lang: en-GB
author: "Lê Quốc Khôi"
date: 2026-11-02
updated: 2026-11-02
topics: ["Corporate finance"]
tags: ["BCG matrix", "Capital allocation", "Portfolio strategy", "Divestment"]
---

# Unilever and ice cream: the BCG matrix as a capital tool, not a verdict

## Key takeaways

- On 19 March 2024 Unilever announced about 7,500 job cuts and a separation of its ice-cream business, due to finish by the end of 2025. The unit had global sales of €7.9bn the year before (BBC News, 2024).
- The reasons given were that ice cream has less in common with Unilever's other lines, because it needs a frozen supply chain and is more seasonal (BBC News, 2024). The articles give neither the unit's relative share nor its market growth, so it cannot be placed in any cell.
- Henderson built the matrix to balance cash flow: slow-growth, high-share businesses throw off cash, and high-growth ones need it (Henderson, 1970).
- Separating a unit does not make it a "dog". Unilever's chairman said the separation would create a world-leading ice-cream business with strong growth prospects as a standalone company (BBC News, 2024).
- Four traps: where you draw the market boundary, a growth threshold you choose yourself, high share not guaranteeing high profit, and links between units that two axes cannot see.

On 19 March 2024 Unilever said it would separate its ice-cream business from the rest of the group. For a finance director running a four-division company in Vietnam, that headline points back to the budget: where should next year's capital go, and which division is only there out of habit? The [BCG growth–share matrix](https://staging.khoi.co.uk/en/framework/bcg-matrix) was built for that kind of question. It was not built to explain what Unilever decided.

## What did Unilever announce about its ice cream?

According to the BBC, Unilever said it would cut about 7,500 jobs worldwide under a three-year cost-saving plan and, separately, split off its ice-cream business. The separation was to start immediately and be completed by the end of 2025 (BBC News, 2024). This is the [Unilever 2024 case](https://staging.khoi.co.uk/en/case/unilever-restructuring-2024). The two decisions are separate: Sky News reported that Unilever announced the ice-cream spin-off separately, and that the job losses would come mainly from office roles as technology advanced (Sillars, 2024).

The ice-cream division had global sales of €7.9bn in 2023, and the productivity programme aimed to save about €800m over three years (BBC News, 2024; Sillars, 2024).

On the reasoning, Unilever said ice cream has less in common with its other consumer lines because it needs a frozen supply chain and is more seasonal (BBC News, 2024). It wanted a simpler, more focused company, targeting mid-single-digit underlying sales growth and a modest margin improvement after the separation (Sillars, 2024). An analyst at Hargreaves Lansdown said the ice-cream unit had been underperforming (BBC News, 2024; Sillars, 2024). That is an outsider's view, not a measure of market share.

## What question does the BCG matrix answer?

It answers a cash question: in a group with several businesses, where does cash come from and where should it go? Henderson's starting point was that high-growth products need cash to grow, low-growth products should generate surplus cash, and a company needs both at once (Henderson, 1970).

> **Figure:** [BCG growth–share matrix](https://staging.khoi.co.uk/en/framework/bcg-matrix) (redrawn from Henderson, 1970)

A cash cow has high share and low growth. It generates more cash than it needs to hold its share, and Henderson says the surplus should not be pushed back into the same business. A question mark has low share and high growth. It almost always needs far more cash than it earns, and it has value only if it becomes a leader. A star has high share and high growth and nearly always shows a profit. The remaining cell Henderson calls "pets"; the figure uses the more familiar label "dogs". Such a business may show an accounting profit, but it must reinvest all of it to hold its share, and Henderson calls it essentially worthless except in liquidation (Henderson, 1970).

## What can the BCG matrix say about Unilever's decision?

Very little. To place ice cream in a cell you need the growth rate of the ice-cream market and Unilever's relative share in it. Neither article gives either figure (BBC News, 2024; Sillars, 2024), so any claim that "ice cream is a dog" or "ice cream is a cow" is guesswork.

The reasons that were given are of another kind. A frozen supply chain and seasonality are about how much the unit shares operationally with the rest of the group (BBC News, 2024). Weighing options to maximise returns for shareholders, and aiming for a simpler, more focused company, are about the shape of the group after the split (Sillars, 2024). The matrix has no axis for any of this.

The most telling detail is the chairman Ian Meakins's statement: the separation would create a world-leading ice-cream business with strong growth prospects as a standalone business (BBC News, 2024). That is management talking about its own decision, not an independent assessment. Even so, the public message does not put ice cream in the exit cell. [Divestment](https://staging.khoi.co.uk/en/glossary#divestment) by demerger or sale is a decision about group structure, not a grade for the unit being separated.

## How should a four-division group use the matrix to decide where the next dong goes?

Do three things in order: fix each unit's market, score the two axes with numbers, then read the cell to pick the capital question. The example below is entirely hypothetical. I made up the division names and every figure to show the method; none of it comes from a real company or statistic.

**A hypothetical example.** A group in the Mekong Delta has four divisions and VND 50bn to invest next year. Relative share here means the division's sales divided by the sales of the largest rival in the same market. The thresholds are 10 per cent a year for market growth and 1.0 for relative share.

| Division | Market growth a year | Division sales | Largest rival | Relative share | Cell |
|:--|--:|--:|--:|--:|:--|
| Own-brand packaged rice | 3% | VND 240bn | VND 160bn | 1.5 | Cash cow |
| Cold storage and chilled delivery | 18% | VND 60bn | VND 200bn | 0.3 | Question mark |
| Ready meals sold online | 15% | VND 90bn | VND 80bn | 1.1 | Star |
| Paper packaging printing | 1% | VND 30bn | VND 150bn | 0.2 | Dog |

> **Figure:** (matrix) Made-up figures for illustration only, not market data.

```json
{
  "type": "matrix",
  "x": { "label": "Relative share", "low": "Below 1.0", "high": "1.0 or above" },
  "y": { "label": "Market growth", "low": "Below 10% a year", "high": "10% a year or above" },
  "cells": [
    { "key": "question-marks", "title": "Question mark: cold storage", "body": "Growth 18%, relative share 0.3. Needs large funding, or a niche it can lead.", "risk": "high" },
    { "key": "stars", "title": "Star: ready meals", "body": "Growth 15%, relative share 1.1. First call on new capital, but track the unit's own cash flow.", "risk": "mid" },
    { "key": "dogs", "title": "Dog: printing", "body": "Growth 1%, relative share 0.2. No new capital until its role inside the group is checked.", "risk": "high" },
    { "key": "cash-cows", "title": "Cash cow: packaged rice", "body": "Growth 3%, relative share 1.5. Hold share and move the surplus cash elsewhere.", "risk": "low" }
  ]
}
```

Reading the table as capital questions:

- Packaged rice is the cash cow. Fund only what it takes to defend share, then measure the [free cash flow](https://staging.khoi.co.uk/en/glossary#free-cash-flow) left over to see how much can move. Henderson argues that if the rate of return exceeds the growth rate, this surplus cannot be reinvested in the same business indefinitely without depressing returns (Henderson, 1970).
- Ready meals are the star. Give them most of the new money, say VND 30bn of the assumed 50bn, but work out the unit's own cash flow, since a star may not fund all of its growth itself (Henderson, 1970).
- Cold storage is the question mark. Henderson writes that question marks need large cash inputs to buy share and are a liability unless they become leaders (Henderson, 1970). The practical test is whether any niche is narrow enough for the remaining funds, say VND 20bn, to put the unit in front. If none is, stop.
- Printing is a dog on the matrix, but do not sell it yet. See the fourth trap below.

## Which traps make a placement misleading?

First trap: relative share only means something once the market is fixed. If cold storage is the whole national market, the figure is 0.3. Suppose instead the market is cold storage for seafood in the Mekong Delta, and VND 45bn of the division's VND 60bn sales come from that niche. The largest rival in the niche has VND 38bn and the niche also grows 18 per cent. The figure is then about 1.2 and the unit moves from question mark to star. The unit has not changed; only the boundary has. That is a [segmentation](https://staging.khoi.co.uk/en/glossary#market-segmentation) decision.

Second trap: the growth threshold is yours to choose. At 10 per cent, ready meals (15 per cent) are a star. Raise the line to 20 per cent and they become a cash cow, purely because the line moved.

Third trap: high share does not guarantee high profit. Henderson holds that margins and share go together, citing common observation and the experience curve effect (Henderson, 1970). This site's framework entry cautions that the link is not automatic [BCG growth–share matrix](https://staging.khoi.co.uk/en/framework/bcg-matrix). To know whether a unit earns its keep, read the unit's own accounts and compare its return with the [opportunity cost of capital](https://staging.khoi.co.uk/en/glossary#opportunity-cost-of-capital).

Fourth trap, outside the two axes: the matrix cannot see how units depend on each other. Suppose the printing division makes the boxes for packaged rice at below-market prices. Selling it could push up the cow's costs, so the "dog" needs one more calculation: the sale price against the cost of buying boxes outside. That too is hypothetical.

## Going deeper: where does the matrix come from, and where are its limits?

Bruce Henderson published the idea in "The Product Portfolio" in 1970, in BCG Perspectives, the publication of The Boston Consulting Group (BCG) (Henderson, 1970). The article sets out four rules of cash flow: margins and cash generated depend on market share; growth requires cash to finance added assets; high share must be earned or bought, and buying costs extra investment; and no market grows for ever, so the payoff from growth comes only when growth slows (Henderson, 1970). Cash earned in one business and spent in another inside the same group is how an [internal capital market](https://staging.khoi.co.uk/en/glossary#internal-capital-market) works.

The weaknesses sit in the assumptions. The first rule, that high share goes with high margin, rests on the experience curve effect in Henderson's account (Henderson, 1970). It is an assumption of the model, not a law, and each industry has to test it. The text of Henderson's article gives no numerical threshold for "high" and "low", and none of this post's three sources contains an empirical test of the matrix. That is why the matrix works as the opening of a discussion about capital, and not as a substitute for judgement at the end of it.

## Frequently asked questions

### Can a small company use the BCG matrix?

Yes, if it has at least two business lines and cash that has to be shared between them. Henderson wrote about the product portfolio of a diversified company (Henderson, 1970). A firm with a single product has nothing to compare.

### Does a unit stay in one cell for ever?

No. Henderson says every product eventually becomes either a cash cow or a pet, and a star that stays a leader turns into a cash cow once growth slows (Henderson, 1970). A cell is therefore a snapshot, so rescore when the market slows or the market boundary moves.

### How is a demerger different from selling a business?

In a demerger, existing shareholders receive shares in a newly listed company (BBC News, 2024). In a direct sale, a buyer pays the group. The BBC reported a demerger as the most likely route, but Unilever said it had not ruled out others, such as a direct sale (BBC News, 2024).

### Does the BCG matrix replace return calculations for a project?

No. It tells you which question to ask about each unit's capital. The actual sums still go through capital budgeting and are compared with the opportunity cost of capital.

## References

- BBC News (2024) *Unilever to cut 7,500 jobs and spin off ice cream business*. BBC News. Available at: https://www.bbc.com/news/business-68595206.
- Henderson, B.D. (1970) *The product portfolio*. BCG Perspectives (The Boston Consulting Group). Available at: https://www.bcg.com/publications/1970/strategy-the-product-portfolio.
- Sillars, J. (2024) *Marmite maker Unilever to cut 7,500 jobs*. Sky News. Available at: https://news.sky.com/story/marmite-maker-unilever-to-cut-7-500-jobs-13097884.

---

Canonical HTML version: https://staging.khoi.co.uk/en/blog/unilever-tach-kem-ma-tran-bcg