GE-McKinsey Matrix: How to Prioritize Business Units

When you run a multi-unit business, every dollar invested in one division is a dollar not invested in another. The GE-McKinsey matrix gives executives a structured way to make those trade-offs without relying on gut feel.

What is the GE-McKinsey Matrix?

The GE-McKinsey Matrix is a 9-cell portfolio framework that helps companies evaluate and prioritize their strategic business units (SBUs) by plotting two dimensions: industry attractiveness and competitive strength. Each unit lands in one of three zones, Invest, Selective, or Harvest, which drives the investment decision.

Key Facts

• The GE-McKinsey Matrix was developed in the 1970s by McKinsey & Company for General Electric to help GE prioritize its ~150 strategic business units (McKinsey Quarterly, 2008).
• About 47% of Fortune 500 companies use a 9-box portfolio framework like GE-McKinsey or BCG to guide investment decisions across business units (Strategic Management Society, 2024).
• GE used the matrix to divest more than 100 business units between 1981 and 2001 under Jack Welch's "Be #1 or #2" mandate, a direct application of the harvest/divest zone logic (Harvard Business Review, 2014).

GE commissioned McKinsey in the early 1970s to solve a problem that every diversified conglomerate knows: how do you allocate capital rationally across 150 business units spanning jet engines, appliances, financial services, and broadcasting? The BCG Matrix existed, but GE's leaders found its 2x2 simplicity too coarse for complex portfolio decisions. McKinsey's solution was a 3x3 grid with weighted scoring, giving strategists nine zones instead of four and a systematic way to quantify factors that BCG left to guesswork.

The result became one of the most widely taught portfolio tools in business schools and one of the most practically used frameworks by corporate strategy teams.

The Two Axes: Industry Attractiveness and Competitive Strength

Every business unit's position on the matrix comes from two composite scores. Getting these right is where most of the analytical work happens.

Industry attractiveness measures how appealing the market itself is, independent of who is competing in it. A high-attractiveness industry grows fast, has strong profit margins, faces limited regulation, and isn't dominated by a few price-setting giants. A low-attractiveness industry might be shrinking, commoditized, or structurally unprofitable. This axis is partly about external opportunity and partly about whether the structural economics of the market reward participants.

Competitive strength measures how well a specific business unit is positioned to win in its industry. Strong competitive position means the unit has superior market share, a recognized brand, lower cost structures than rivals, proprietary technology, or distribution advantages that are hard to replicate. Weak competitive position means the unit is fighting uphill: it's a follower, not a leader, and rivals can undercut or outpace it on most dimensions.

Both axes are scored on a scale, typically 1 to 5 or 1 to 10, using a set of weighted factors. The weighting step is critical: a factor like market size matters, but it probably matters less than market growth rate when you're making a forward-looking investment call.

The 9 Cells and 3 Investment Zones

The 3x3 grid creates nine cells, but strategically they collapse into three zones. Think of the matrix as a diagonal dividing line running from top-left to bottom-right. Everything above and to the left of that diagonal is where you want to invest. Everything on the diagonal warrants careful selectivity. Everything below and to the right should be harvested or divested.

Invest/Grow (Top-Left 3 Cells)

Business units in the invest zone sit in attractive industries where the company already has strong competitive position. These are your best units: the market wants to grow and you're well-placed to capture that growth. Capital, management attention, and talent should flow here. Jack Welch's application of this logic at GE was almost literal: if a unit wasn't first or second in its market, it got fixed, sold, or closed.

Don't mistake "invest" for "no scrutiny." The best-performing units still need clear growth plans, accountability metrics, and ROI targets. Unrestricted capital into a strong unit without a growth roadmap is just waste at a better address.

Selective (Diagonal 3 Cells)

The three diagonal cells represent genuine ambiguity. A unit might operate in an attractive market but hold a weak competitive position. Or it might be the market leader in an industry that's no longer growing. Neither is automatically worth doubling down on or walking away from.

The strategic call here requires deeper investigation. Can you improve the unit's competitive position through investment in capability, brand, or distribution? Is the market forecast accurate or too pessimistic? Are there strategic synergies with other units that the standalone score doesn't capture? Selective units demand a hypothesis: you either have a credible plan to move this unit into the invest zone, or you accept it will eventually slide toward harvest.

Harvest/Divest (Bottom-Right 3 Cells)

Units in the harvest zone sit in unattractive industries where the company's competitive position is also weak. These are the hardest conversations because they often involve legacy products, loyal long-tenured teams, or businesses that were profitable in previous decades but have since been commoditized.

The strategic logic is clear: stop allocating growth capital here. If the unit generates cash, harvest it (run it efficiently, extract margin, don't reinvest for expansion). If it's a cash drain, consider divesting before it destroys further value. This isn't defeatism; it's capital discipline. Every dollar freed from a harvest unit can fund an invest unit's growth plan.

How to Score Industry Attractiveness and Competitive Strength

The power of the GE-McKinsey Matrix over simpler tools like the BCG Matrix is that you score each axis using multiple weighted factors, not a single proxy metric. Here's how each axis is typically constructed.

Industry Attractiveness Factors

Competitive Strength Factors

Weights should add to 100% per axis and should be agreed upon by the leadership team before scoring begins. Changing weights after seeing scores introduces bias.

How to Build a GE-McKinsey Matrix: Step by Step

Step 1: Identify Your Business Units

Define the units you're evaluating. These should be genuinely distinct: separate customer segments, revenue streams, cost structures, and competitive environments. Lumping dissimilar products into one "unit" muddies the analysis. Think of each unit as a separate strategic entity that could theoretically be evaluated as a standalone business.

Step 2: Define the Scoring Factors

Select 6-8 factors for each axis. Use the tables above as a starting point, but customize for your industry. A pharmaceutical company might weight regulatory factors higher; a consumer brand might weight distribution reach more heavily. Run your factor list by multiple leaders before finalizing, because the selection itself surfaces strategic assumptions that are worth debating.

Step 3: Weight the Factors

Assign weights to each factor on both axes. Weights should reflect what actually drives success in your markets, not what's easiest to measure. Total weights per axis must sum to 100%. Do this step before scoring any specific units; otherwise you risk reverse-engineering weights to justify a preferred answer.

Step 4: Score Each Business Unit

Score each unit on each factor using a consistent scale (1 = very weak, 5 = very strong for competitive strength; 1 = very unattractive, 5 = very attractive for industry attractiveness). Then multiply each score by its weight and sum the weighted scores to get a composite score per axis. Involve people who have direct knowledge of the unit and its market: finance, product, and sales leadership typically.

Step 5: Plot the Matrix

Map each unit onto the 3x3 grid using its composite scores. If you're using a 5-point scale, scores 1-2 are low, 2-3 are medium, and 3-5 are high. Use circle size to represent revenue or assets, which adds a third dimension showing how much is at stake in each cell. This is where the conversation starts, not ends.

Step 6: Set Strategy per Zone

Once units are plotted, assign one of three strategic postures: invest for growth, selective attention with a specific improvement hypothesis, or harvest and redeploy. Document the rationale for each, including what metrics would change the conclusion. The matrix should drive resource allocation decisions, budget conversations, and portfolio review cadences.

GE-McKinsey Matrix Examples

Here's how a diversified technology company might apply the matrix across four of its units.

The Cloud SaaS Platform gets the bulk of next year's growth investment. The Legacy CRM gets a maintenance budget and a timeline. Management attention follows the same logic: the strongest leaders get assigned to Invest units, not spread evenly.

This is also where the balanced scorecard becomes a useful companion: once investment decisions are made, the scorecard translates them into specific KPIs and operational targets for each unit.

GE-McKinsey vs BCG Matrix

Both frameworks evaluate business portfolios, but they differ in meaningful ways. Knowing when to use which one saves a lot of wasted analysis.

For most large enterprises with genuine portfolio complexity, the GE-McKinsey framework delivers more actionable output. The BCG Matrix works well as a rapid diagnostic or when data is limited.

Limitations and Criticisms

No framework survives contact with reality perfectly. The GE-McKinsey Matrix has real limitations worth knowing before you apply it.

• Subjective scoring. Weighted factors are only as good as the people scoring them. Teams tend to inflate scores for units they manage or favor. Build in red-team review or external benchmarking to counter this.
• Snapshot in time. The matrix reflects current and near-term conditions. It doesn't model disruptive shifts well. A unit in the invest zone today can move to harvest if a new entrant rewrites the competitive landscape. Revisit annually at minimum.
• Ignores synergies. Two units might each score in the harvest zone independently but together create capabilities neither has alone. Cross-unit dependencies require qualitative overlay that the matrix doesn't capture on its own.
• Data-hungry. Building an honest, rigorous matrix requires competitive intelligence, market sizing data, and internal financials that aren't always available or reliable, especially for smaller units or emerging markets.
• Can bias toward incumbency. Strong competitive position rewards existing scale and market share. Disruptive or nascent units within a portfolio may look weaker than they are because their strength lies in future potential, not current metrics.

Used alongside other tools like PESTEL analysis and Porter's Five Forces, the matrix becomes significantly more reliable. The external environment analysis from PESTEL feeds directly into the industry attractiveness scoring; Five Forces provides the competitive intensity factor.

Best Practices

• Agree on weights before scoring. Locking in factor weights before any units are scored prevents post-hoc rationalization.
• Use circle size for revenue. Plot units as circles proportional to their revenue or asset base. A small circle in the harvest zone is a different conversation than a large one.
• Run the exercise annually. Market conditions change. A unit that was selective last year may have moved into invest or harvest territory. Treat the matrix as a living document, not a one-time project.
• Separate the analysis from the decision. The matrix informs strategy; it doesn't replace judgment. Bring qualitative context, particularly around synergies, talent, and strategic options, into the room before finalizing recommendations.
• Involve multiple functions. Finance, product, and commercial leaders each have information the others don't. Cross-functional scoring panels produce more accurate results than top-down executive estimates.
• Challenge comfortable conclusions. If every unit lands in the middle of the grid, your scoring criteria are probably too narrow. Push for genuine differentiation in the scores.
• Connect it to resource allocation. The matrix has no value unless it drives actual budget and headcount decisions. Tie each zone recommendation to a specific capital allocation or divestiture plan.
• Pair it with the Ansoff Matrix. Once you know which units to invest in, Ansoff helps you decide whether to grow through market penetration, new markets, new products, or diversification.

Frequently Asked Questions

What's the main difference between the GE-McKinsey Matrix and the BCG Matrix?

The BCG Matrix uses two single metrics: market growth rate and relative market share. The GE-McKinsey Matrix replaces those with composite scores built from 6-8 weighted factors per axis. This makes GE-McKinsey more accurate and more useful for complex portfolios, but also more time-intensive to build. Use BCG for a quick scan; use GE-McKinsey when the stakes are high enough to warrant rigorous analysis.

How often should a company update its GE-McKinsey Matrix?

Most strategy teams update the matrix annually as part of the strategic planning cycle. But significant events, such as a new market entrant, a major acquisition, or a sharp shift in macroeconomic conditions, should trigger an out-of-cycle review. The PESTEL analysis and Porter's Five Forces frameworks can help you spot when conditions have shifted enough to warrant a matrix refresh.

Can a startup or early-stage company use the GE-McKinsey Matrix?

Technically yes, but practically it's often premature. The matrix works best when a company has multiple distinct business units with enough operating history to score on profitability, market share, and brand strength. A startup with one or two nascent products lacks the comparative data the framework needs. The Business Model Canvas or Value Proposition Canvas are more appropriate tools at that stage.

Who should score the factors in the matrix?

Scoring should involve a cross-functional team: typically the CFO or FP&A lead for financial metrics, business unit leaders for operational and competitive data, and a central strategy function for market intelligence and external benchmarking. Avoid letting individual unit leaders score their own units without external review; self-assessment bias is real and predictable.

Can the GE-McKinsey Matrix be used to evaluate product lines rather than business units?

Yes. The framework adapts well to product portfolio decisions as long as each product has distinct economics, market dynamics, and competitive positioning. The logic is identical: score each product on industry attractiveness and competitive strength, plot them, and set investment priorities accordingly. Just make sure the "industry" definition is scoped appropriately for a product-level analysis rather than a business-unit-level one.

Where to Go From Here

The GE-McKinsey Matrix gives you a portfolio map: which units deserve investment, which need a hypothesis for improvement, and which should be harvested. But a map is only useful if you follow it.

The next layer of work is translating matrix outputs into strategy execution. That means setting OKRs for invest units that reflect aggressive growth targets, building SMART objectives for selective units around the specific capabilities they need to develop, and running value chain analysis to find where harvest units can cut cost while maintaining margin.

Portfolio strategy without execution is just analysis. The matrix tells you where to point the ship. The management system you build around it determines whether you actually get there.