SDA Professor of Strategic and Entrepreneurial Management
The BCG matrix and modified portfolio planning tools have for a long period of time occupied a fundamental position in corporate planning departments as well as in business school courses. With this contribution, I attempt to show that the BCG matrix has a number of flaws and its application is rather problematic. The Boston Consulting Group’s (BCG) growth share matrix – and its close relatives (most prominently the GE/McKinsey matrix) - is one of the best known and persistent tools in strategic management. At the height of its success between 1972 and 1982, the BCG matrix was used by around 45% of the Fortune 500 companies. In 1975, the prominent strategy scholar Peter Lorange asserted that the growth share matrix has become the common method of corporate planning.
The apparent simplicity of having reduced a complex decision problem to a two dimensional matrix was of intuitive appeal. The central assumption was based on academic research (i.e., the PIMS study) and managers located in corporate headquarters were able to show their value added for their businesses. The proliferation of portfolio planning as a resource allocation tool was accelerated by the BCG consultants and their competitors like McKinsey, and Arthur D. Little which developed similar matrixes together with their clients.
Portfolio matrixes explained
The BCG matrix explicitly recognizes that a diversified company is a portfolio of businesses, each of which should make a distinct contribution to the overall corporate performance and should be managed accordingly. Portfolio planning has its roots in the late 60s as firms started to becoming larger and more diversified. Corporate headquarters were created with the aim to add value to their collection of businesses by allocating funds, creating synergies or directly influencing their activities. A central concern of corporate managers was to allocate funds in a strategic way and thereby to ensure that the company balances the cash needs of its business units. Alternative resource allocation methods such as basing investment decisions on track records or the complex prediction of discounted project cash flows did not seem to entirely satisfy corporate managers. The BCG matrix on the other hand helps to allocate limited resources based on one single graph. As such, it serves as a simplifier that links a single market dimension (i.e., market growth) to a single company dimension (i.e., relative market share).
The BCG claim was that the four different business categories in the different squares of the matrix have indeed different tendencies to generate or consume cash and therefore need to be treated accordingly:
- “Question Mark” businesses (high growth and low relative market share) have high cash needs and firms should therefore do whatever is necessary to increase market share or divest quickly.
- “Star” businesses (high growth and high relative market share) are frequently roughly in balance on net cash flow and may exceptionally need more cash than what is generated by their own business. The main focus of star businesses is to protect the market share, hence to get a bigger portion of the market growth than its competitors. Star businesses will ultimately turn into cash-cows, assuming that markets cannot grow forever or become dogs if they fail to protect the market share.
- “Cash Cow” businesses (low growth and high relative market share) are characterized by high profit and cash generation. The remaining cash after covering costs to run the business and to protect the share in a mature market should be redistributed to other businesses.
- “Dog” businesses (low growth and low relative market share) generate often poor profits and cash needs are frequently higher than the cash that is generated. To improve the overall performance, firms should minimize the proportion of their assets that remain in this category by focusing on a specialized segment, harvesting by cutting costs and maximizing cash flow by divestment or liquidation.
The pitfalls of portfolio matrices
But most portfolio matrices have several flaws. Now, let’s assume for a moment that you are running a business unit that has been classified as a “cash cow”, because of its limited market attractiveness (in terms of market growth, market size and intensity of competition) and your dominant market position (in terms of relative market share, technological strength and brand name). Following the portfolio logic, your CEO has decided to “milk” your business and would not let you invest substantially. What are the arguments you could use to convince your CEO that not investing in your business would be a mistake? What are the pitfalls of portfolio matrices?
Here are a few:
- Importance of market share: not in all businesses one would find a relationship between market share and profitability. The portfolio matrix logic usually gives too much emphasis on cost leadership rather than differentiation as a source of competitive advantage. What is the role of market share for Hermès, Ferrari or the full-time MBA program of SDA Bocconi?
- Measurement problems: relative market share depends on how the market is defined – and that is rather subjective. Narrowing the definition of the reference market can turn a question mark into a star. Both axes, relative market share and market growth, depend on mangers’ subjective interpretation of what the relevant market is. Broad versus narrow definition of markets reflect different views on the long-term development of the business. In numerous markets, it is difficult to gather data about the total market size, its growth rate and the share of the biggest competitor, not to mention the measurement and forecasting of product life cycles. The impact of re-launches, technological innovation and face-liftings on product life cycles is highly uncertain and further complicates growth rate predictions. The measurement problems may lead to yield different recommendations for the same situation
- Assumptions on product-life-cycle: the chronological sequence of questions marks (start-up) becoming stars (growth), then cash-cows (maturity) and eventually dogs (turnaround or decline/sale) has not proven to be true for all businesses. Think about Coca Cola, or Aspirin. They have been star businesses for a very long period of time. But even poor dogs (nota bene: they have been initially labelled “pets”) often compete against lazy cash-cows and dying dogs and therefore provide stable cash flows – often more cash than the question mark businesses burn. Hence, the generic suggestion “milk” or “divest” might be misleading.
- Over simplistic: the BCG model’s use of only two dimensions (growth and share) to assess the competitive position has been partly compensated by including other elements to assess market attractiveness and relative strengths. However, it remains a relatively simple representation of realty. While this partly explains the attractiveness of the model for managers, we need beware of the pitfall to blindly follow the model’s indications.
- Hard to implement generic strategies: when corporate level has allocated a certain amount of money to a single business unit, the problem of allocating the money to single projects still remains. Another major implementation problem is the difficulty one faces to unequivocally operationalize strategic guidelines such as “milk for cash”, “harvest”, or “selectively invest”. Although intuitively appealing, those normative strategies need to be contextualized and refined. The BCG logic implies that firms have the ability to shift resources from one business to another. In capital intensive and politically-sensitive businesses, it might be costly to liquidate or divest a dog business. Barriers to exit due to legal or political constrains may complicate the implementation of the BCG guidelines. Even though there’s usually a private equity fund or emerging market MNE willing to buy a dog business if the price is right, many firms are not good at divesting businesses and could learn from Richard Branson who has done a brilliant job of exiting dog businesses by floating them or selling them (in whole or part) to other companies e.g. the airlines Virgin Atlantic, Virgin america, Virgin austrlia, the telecom businesses Virgin mobile and Virgin Media, and the healthclubs, Virgin Active.
- Synergies are not considered: portfolio matrices focus too much on balancing cash flows rather than other interdependencies. For example, not investing in cash cows or divesting dogs may imply that star businesses lose access to resources and capabilities that are essential for their development. This phenomenon has been labelled as “phantom limb” effect of business divestment. Similar to the phantom pain patients feel for years after they have lost a limb, missing links between business units hurt company performance. Some matrices remedy by including “synergies” in the list of factors they consider when assessing relative strengths. But the basic problem remains: we are using a portfolio approach for companies that are simply not a portfolio of (independent) businesses.
- Backward looking: The positioning of all businesses according to their relative market share and their market growth results in a one spot representation of the business portfolio. The goal of management should be to have a balanced portfolio to ensure an optimization of the cash streams and with them the overall performance of the company. Hence, the sustainable growth rate of a company is seen as a function of its portfolio of cash-generating businesses and cash-using businesses. The experience curve effect explains why high market share businesses generate more cash than they can meaningfully redeploy and therefore create opportunities for new businesses in growing markets. However, the data on recent market growth in particular is not a good indicator of what market growth is likely to be in the future. For the producers of luxury goods, their “star” markets China and Brazil are not looking so attractive in 2015.
- Perception biases: We know simple labels have a huge influence on how people assess objects (e.g., if someone has been labelled “a genius”, everything that person does tends to confirm that perception in our minds). The BCG matrix attaches very broad (and imprecise) labels to business units that will over and over influence how those units are perceived within the organization. It will also influence how top management filters the information about those units. Make the test: take an investment opportunity and see whether this is assessed differently whether it is coming from a star or from a dog business simply because everything associated with the dog label will be perceived more negatively in the mind of a manager.
Don’t get me wrong – I think that a well-constructed portfolio matrix can support resource allocation decisions. But as most management tools, it should only be a basis for discussion that is complemented by other approaches, such as the valuation of concrete investment proposals (DCF method), or the evaluation of past success of applicant. So when your management team next time tries to “milk” your cash-cow business or “divest” your dog business, use some of the arguments listed above to get attention for your alternative investment proposal.
*I would like to thank David Bardolet, Robert Grant and Francesco Saviozzi for their comments.Clicca qui per la versione stampabile