Evaluating the BCG Matrix
Does the famous BCG Matrix (and arguably the very similar GE-McKinsey Matrix) help formulate a good strategy?
The BCG (or a more informative name being the “growth-share”) matrix is a famous staple taught in Strategic Management courses at business schools.
It seems intuitive and commonsensical, and is likely useful and true to some extent, but what are the potential challenges and issues in using it?
Some practical issues when using the matrix
A common saying is “What gets measured (wrongly) gets managed (wrongly).” In terms of practical issues:
(1) How do we measure market share and growth? Annually? Now or dynamically over time (future projections)? More importantly, how to define the market (and what is not — i.e. its boundaries)[i]? Define too narrowly and we get a huge market share, and vice versa.
(2) Which quadrant does a business fall in? Market growth and share are continuous, yet we sort them into high or low[ii]. Even if we have some cut-off/threshold[iii], two products on just on either side do not really differ in a meaningful sense[iv]. Should we really apply significantly different strategies?
(3) The growth-share matrix only considers two, though relatively easily measured, metrics. However, are they right, or enough?[v]. A similar matrix, the GE-McKinsey-Matrix (see image above) aims to broaden the considerations, but this results in the opposite problem: Industry attractiveness and business unit strength are complex and cannot be measured directly. What components (and their weightages) should comprise them?
(4) It is one thing to recommend “invest”, “milk”, “analyze”, and “divest”, but the devil is in the details. Throwing money at something is not the same as deciding to where and how to invest (marketing, talent, etc?). How exactly are we to “analyze” a Question-Mark? The fact there are 2 different paths for a product’s lifecycle in BCG-Matrix suggests that it itself does not allow us to make predictions[vi].
(5) The strategies may not be appropriate for different time horizons, e.g., Are pets really worthless — or could a market that is not ready now change in the future?
How valid are the matrices?
Beyond measurement and identification problems, is the growth-share matrix, and its recommendations, valid in the first place?
Conceptually, if we take market share as given, are we neglecting strategies to increase it?
Also, will following the growth-share matrix result in profits or competitive advantage? Indeed, the empirical evidence is mixed. While Hambrick et al. (1982) found that, for industrial businesses, the growth-share matrix’s cash flow predictions were corroborated, the average Dog had a positive cash flow, more than the average needs of Question-Marks, implying that they need not be liquidated.
Insofar as BCG is meant to be useful for managers, Capon et al. (1987) [vii] showed, in a field study, that firms using BCG reported lower ROC[viii]. Armstrong and Brodie (1994), in a lab experiment, reported 87%[ix] and 64%[x] of those using and exposed to BCG selected the unprofitable investment respectively[xi].
Conclusion
Ultimately, the growth-share matrix is just a( potentially useful) tool to organize our thoughts and ensure we do not miss important considerations (internal and external).
However, tools simplify[xii], perhaps simplifying away aspects that are relevant (e.g. boundary conditions, contingency factors), and even introduce biases (e.g. mental accounting[xiii] and emotional manipulation[xiv]).
Therefore, we must not be tempted to seek safe solutions in any general framework (heuristic), no matter how supposedly logical and authoritative[xv]. The growth-share matrix tells us what is, but not necessarily what should be.
While they may help organize our thoughts and keep in mind important considerations, ultimately, there is also no substitute for fundamental and specific strategic analysis, based on hard work, critical thinking, data and experimentation, and making difficult choices[xvi].
Sources and References
Armstrong, J. Scott, and Roderick J. Brodie. “Effects of portfolio planning methods on decision making: Experimental results.” International Journal of Research in Marketing 11, no. 1 (1994): 73–84.
Hambrick, Donald C., Ian C. MacMillan, and Diana L. Day. “Strategic attributes and performance in the BCG matrix — A PIMS-based analysis of industrial product businesses.” Academy of Management Journal 25, no. 3 (1982): 510–531.
King, Andrew A., and Baljir Baatartogtokh. “How useful is the theory of disruptive innovation?.” MIT Sloan Management Review 57, no. 1 (2015): 77.
Madsen, Dag Øivind. “Not dead yet: the rise, fall and persistence of the BCG Matrix.” Problems and Perspectives in Management 15, no. 1 (2017): 19–34.
Reeves, Martin, Sandy Moose, and Thijs Venema. “BCG classics revisited: The growth share matrix.” BCG Perspectives (2014).
[i] A related and notable mistake in identifying the market was made by Christensen when applying his disruptive innovation theory to the iPhone, and from it, predicting its failure. He admitted that he had “framed the problem wrongly”, viewing Apple “as a late entrant into the mobile phone business” whereas it should be the “computer in your pocket business” (2020). However, this begets the question, if we can always redefine the application of the theory ex post, to show that it was right “all along”, but it does not allow for ex ante predictions, then what is the point of a theory or framework?
[ii] The GE-McKinsey Matrix improves the growth-share matrix by having an intermediate category, but the same critique applies
[iii] This likely varies for different industries, and even then, may be subjective (no true benchmark other than itself?)
[iv] cf. regression discontinuity
[v] BCG (2014) themselves acknowledge that there are other drivers, though they contend these are new drivers brought up a changing environment, such as a “new measure of competitiveness to replace its horizontal axis now that market share is no longer a strong predictor of performance” and “the ability to adapt to changing circumstances, or to shape them”.
[vi] Which is what theories should aim for — providing useful advice for individuals and organizations (Bazerman, 2005)
[vii] cited by Armstrong & Brodie (1994)
[viii] One may argue reverse causality — that unprofitable firms turn to the BCG matrix for guidance
[ix] Compared to only 15.3% of those using NPV calculations
[x] Compared to 44.7% in the control group; the difference was statistically significant at p < 0.01
[xi] Note that the lab experiment specifically had the BCG matrix implications leading to an unprofitable decision (despite other, sufficient information to the contrary); the question was whether it would be used blindly; still, Hambrick et al., (1982) found that many Dogs have high profits, implying that BCG might not result in profit-maximization
[xii] Krogerus & Tschappeler, 2015
[xiii] Categorizing a business into a quadrant may lead managers to use the capital allocated differently (Armstrong & Brodie, 1994)
[xiv] Categorizing a business into a quadrant with a highly charged label (“Dogs”; “Stars”) may emotionally influence how such businesses are viewed (Armstrong & Brodie, 1994)
[xv] Madsen (2017) provides an explanation for the rise and persistence of the BCG Matrix through the lens of management fashion theory.
[xvi] cf. King & Baatartogtokh, 2015, in the context of disruptive innovation theory
