Strategy & competition
GE–McKinsey Nine-Box Matrix
A nine-cell portfolio grid that rates each business unit on industry attractiveness and competitive strength, then prescribes invest and grow, selectivity, or harvest and divest, so that corporate capital follows structural merit rather than internal politics.
Also known as GE matrix, McKinsey nine-box, Industry attractiveness–business strength matrix, GE multifactorial analysis. First set out by McKinsey & Company, for General Electric in 1971; the primary source is cited in full below.
Where this is contested
Attribution is institutional rather than personal: the matrix was developed jointly by McKinsey consultants and GE's corporate planning staff in the early 1970s as a successor to BCG's growth-share matrix, and no single originator or definitive first publication exists; sources date it anywhere from 1970 to the mid-1970s.
- Format
- Mapping
- Level
- Corporate · Business unit
- Best for
- Allocate resources · Prioritise · Evaluate options
- Decision stage
- Diagnose · Decide
- Difficulty
- Advanced
- Time to apply
- Two to four weeks for a serious corporate pass with agreed criteria and real market data; a day for a workshop first cut.
Plate · The model
The components
Protect position (high attractiveness, strong unit)
The portfolio's crown jewels. Invest to grow at the maximum rate the unit can digest, and concentrate effort on maintaining the strengths that put it here.
Signals of strength
Leads an attractive, growing market · Returns comfortably above the cost of capital · Growth constrained by capacity rather than demand
Invest to build (high attractiveness, average unit)
An attractive market where the unit is not yet a leader. Challenge for leadership: build selectively on strengths and reinforce the areas where the unit is vulnerable, accepting that near-term returns will be diluted.
Signals of strength
Attractive industry, second-tier position · Identifiable, fixable capability gaps · Market leaders that are visibly beatable or complacent
Build selectively or exit (high attractiveness, weak unit)
The gambler's cell. Specialise around the unit's limited strengths and seek ways to overcome its weaknesses; if signs of sustainable strength do not appear on a stated timetable, withdraw before the attractive market extracts more capital.
Signals of strength
Attractive market, marginal share and thin economics · One or two defensible niches within reach · Ongoing funding requests with slipping milestones
Build selectively (medium attractiveness, strong unit)
A strong unit in a middling market. Invest heavily in the most attractive segments rather than across the board, build the ability to counter competition, and raise profitability through productivity.
Signals of strength
Clear leadership in a mature or mixed market · Pockets of genuine growth inside a flat industry · Strong cash generation that can fund segment bets
Selectivity and manage for earnings (medium attractiveness, average unit)
The crowded middle. Protect the existing programme, concentrate investment where profitability is good and risk is low, and resist the drift by which average businesses absorb above-average capital.
Signals of strength
Neither leadership nor obvious decline · Adequate but unremarkable returns · Investment cases that rely on hope rather than position
Limited expansion or harvest (medium attractiveness, weak unit)
Look for low-risk ways to expand, and where none exist, minimise investment and rationalise operations. This cell funds nothing ambitious.
Signals of strength
Weak position with no realistic route to strength · Expansion options that are cheap or nothing · Costs that can be taken out faster than revenue declines
Protect and refocus (low attractiveness, strong unit)
A strong position in a poor industry. Manage for current earnings, defend the strengths that produce them, and refocus the unit towards the few segments where the industry still rewards strength.
Signals of strength
Dominant share of a declining or structurally squeezed market · Reliable cash generation with limited reinvestment options · Attractive sub-segments worth defending
Manage for earnings (low attractiveness, average unit)
Protect position in the most profitable segments, upgrade the offer only where it defends earnings, and hold investment to the minimum that keeps the cash flowing.
Signals of strength
Middling position in an unattractive industry · Earnings sustained by pruning rather than growth · No credible case for strategic investment
Divest (low attractiveness, weak unit)
Sell at the time that maximises value, or run down deliberately. Cut fixed costs, avoid further investment, and do not let sentiment or sunk cost argue for one more turnaround plan.
Signals of strength
Weak unit in a poor market · Persistent returns below the cost of capital · More value to a different owner than to the portfolio
When it earns its keep
- You run a multi-business portfolio and need a defensible basis for giving some units capital and refusing others.
- The BCG matrix feels too crude for your businesses, because growth rate and relative share alone miss what actually makes your markets attractive or your units strong.
- An annual strategy review needs a common language for comparing units that operate in unrelated industries.
- You suspect capital is flowing to the loudest divisional voices, and want the allocation argument moved onto explicit, weighted criteria.
And when it doesn't
- You manage a single business. The matrix compares units against each other; with one unit there is nothing to compare.
- You need speed. Scoring attractiveness and strength honestly requires real market data and several weeks; for a rough cut, the BCG matrix is faster.
- The units share deep operational or customer synergies. The matrix scores each unit standalone and will happily recommend divesting a business the rest of the portfolio depends on.
- The scores would be produced by the units being judged. Self-graded homework in this matrix reliably clusters in the investable cells.
How to run it
Before starting, gather the inputs the analysis depends on:
- A clean definition of each strategic business unit, drawn so that units do not overlap markets.
- Agreed, weighted criteria for industry attractiveness: market size, growth, profitability, structural forces, cyclicality, regulation.
- Agreed, weighted criteria for competitive strength: relative share, brand, unit economics, distribution, capability and asset quality.
- Financial data per unit: revenue, margin, capital employed and cash generation, to size the plotted bubbles and sanity-check the scores.
- A corporate view on available capital, because the matrix ranks claims on a finite pot.
- 1
Define the units
List the strategic business units to be compared and fix their boundaries. A unit drawn too broadly averages a strong business with a weak one; drawn too narrowly it flatters niche positions. GE itself applied the grid at several levels, from product line to sector, but each pass must hold the level constant.
- 2
Build the attractiveness axis
Choose the criteria that make an industry worth being in for your company, weight them, and score each unit's industry high, medium or low. This is where the matrix earns its keep over BCG: attractiveness is multi-factor by design, so argue the weights openly before anyone sees the scores.
- 3
Build the strength axis
Do the same for competitive strength: score each unit against the criteria that decide who wins in its industry, anchored in evidence rather than divisional self-assessment. Relative measures beat absolute ones; a 20 per cent margin can be weakness in one industry and dominance in another.
- 4
Plot the portfolio
Place each unit in its cell, drawn as a bubble sized by revenue or capital employed. The picture matters: a portfolio whose revenue sits in the bottom-left corner is a different company from one whose revenue sits top-right, whatever the narrative says.
- 5
Read the three zones and prescribe
The nine cells resolve into three diagonal bands. Units above the diagonal get investment and growth; units on the diagonal get selective investment with tight conditions; units below it are run for cash or sold. Write the prescription per unit and note where it contradicts current spending.
- 6
Stress-test and decide
Re-run the plot with different weights and with next-cycle assumptions, and see which placements survive. Then make the capital calls. A matrix that reshuffles no budget was an away-day, and an expensive one.
Reading the result
A picture of the whole portfolio on two axes, a cell placement and investment prescription for each unit, and a capital allocation argument grounded in explicit criteria. The canonical form is a 3x3 grid with industry attractiveness on one axis and business strength on the other, and that is how this entry is drawn, the nine cells resolving into three diagonal investment zones.
- Read the three diagonal zones before the nine cells: above the diagonal invest and grow, on it be selective, below it harvest or divest.
- Treat cell placements as hypotheses resting on weighted scores. If a placement flips when the weights move a little, the placement is an opinion wearing a chart.
- Compare the prescriptions against actual capital flows. The gap between where the matrix says money should go and where it currently goes is the review's real finding.
A worked example
A regional media group reallocates capital across print, radio, events and digital
Calderdale Media, a family-owned group in the north of England, owns two daily newspapers and a stable of weeklies, two local radio stations, an exhibitions and events division, a growing digital marketing services arm and a legacy recruitment classifieds site. The board has been spreading capital roughly in proportion to revenue, which means print still takes most of it. The strategy director runs the portfolio through the nine-box matrix, with attractiveness and strength criteria weighted and agreed by the board before any unit is scored.
- Invest to build (high attractiveness, average unit)
- The digital marketing services arm lands here. Local businesses are shifting spend to digital and the market is growing at double digits, but the unit is one agency among many with no distinctive capability yet. Prescription: invest to build, with the acquisition of a specialist search marketing shop as the vehicle.
- Build selectively (medium attractiveness, strong unit)
- The events division is the region's leading business exhibitions operator in a market growing modestly. Prescription: fund the two flagship shows and a new awards franchise, and decline the temptation to launch consumer events where the unit has no edge.
- Selectivity and manage for earnings (medium attractiveness, average unit)
- The radio stations hold steady audience share in a flat market being nibbled by streaming. Prescription: maintain, invest only in the breakfast shows that drive advertising yield, and revisit the placement in two years.
- Protect and refocus (low attractiveness, strong unit)
- Print is structurally declining, but the dailies dominate their towns and still generate most of the group's cash. Prescription: manage for earnings, refocus on the profitable weekly titles and the advertisers who still pay to reach their readership, and stop funding circulation-growth initiatives that fight the tide.
- Divest (low attractiveness, weak unit)
- The classifieds site has lost the recruitment market to national job boards and holds no defensible position. Prescription: sell to a consolidator while the remaining traffic still commands a price.
The read. The matrix reverses the group's capital logic: print, the biggest unit, becomes a source of funds rather than a claim on them, and the proceeds from classifieds plus print cash fund the digital acquisition and the events franchises. The board also records the placement it trusts least, digital's attractiveness score, and agrees what evidence would trigger a re-plot. The grid did not make the decision; it made the argument honest.
Pitfalls
- Letting the units being judged supply their own scores. Attractiveness and strength ratings drift towards the investable cells unless scored centrally against evidence.
- Averaging multi-factor scores into the middle cell. A unit that is excellent on two criteria and dire on two is not 'medium'; the offsetting facts are the finding.
- Treating the cell prescriptions as verdicts rather than starting points. The labels are generic strategies from the 1970s, and every placement deserves a why-here and a what-would-change-it.
- Ignoring links between units. The matrix scores businesses standalone, so it can recommend divesting the unit that feeds the others customers, capability or cash.
- Mistaking a snapshot for a forecast. Attractiveness is a view about the future; scoring it from last year's market data plots where the portfolio was, and prescribes accordingly.
What the critics say
Haspeslagh's survey of portfolio planning in large US companies found the benefit lay in the discipline and dialogue of the process, and that mechanical application of grid prescriptions to capital allocation was rare in practice and dangerous where attempted.
Haspeslagh, P. (1982) 'Portfolio Planning: Uses and Limits', Harvard Business Review, January–February 1982.
Empirical comparison of standardised portfolio models showed that a unit's classification changes dramatically with the choice of model, criteria, weights and cut-off points, so the matrix's prescriptions are far less objective than their numerical dress suggests.
Wind, Y., Mahajan, V. and Swire, D. J. (1983) 'An Empirical Comparison of Standardized Portfolio Models', Journal of Marketing, 47(2), pp. 89–99.
The corporate-strategy literature argues that standalone unit attractiveness misses the question that matters at group level: whether this parent adds value to this business. A unit can sit in an investable cell and still be worth more under a different owner.
Goold, M., Campbell, A. and Alexander, M. (1994) Corporate-Level Strategy: Creating Value in the Multibusiness Company. New York: Wiley.
Sources and further reading
- McKinsey & Company (2008) 'Enduring Ideas: The GE–McKinsey nine-box matrix', McKinsey Quarterly, September 2008. ↗
- Hax, A. C. and Majluf, N. S. (1983) 'The Use of the Industry Attractiveness-Business Strength Matrix in Strategic Planning', Interfaces, 13(2), pp. 54–71.
- Haspeslagh, P. (1982) 'Portfolio Planning: Uses and Limits', Harvard Business Review, January–February 1982. ↗
- Wind, Y., Mahajan, V. and Swire, D. J. (1983) 'An Empirical Comparison of Standardized Portfolio Models', Journal of Marketing, 47(2), pp. 89–99.