The myth of sustainable competitive advantage

by

I’ve mentioned to a few people that I was unimpressed with The End of Competitive Advantage by Rita Gunther McGrath. I’ve even be so impolitic as to call it ‘crap’ a few times, after which I’m usually asked why I think this, and I then spend time shooting over references and explaining how I came to my conclusion.

Rather than doing this yet again, I thought I’d compile some notes here which I can then point people at.

The End of Competitive Advantage claims to provide key insights into how business strategy needs to change, moving on from the foundations laid down by Michael Porter all those years ago. A few even called it an ‘important’ book, as they see it as the first proof that sustainable competitive advantage is a thing of the past.

My problem with the book is in three parts:

  1. The book provides insufficient argument and data to prove its thesis.
  2. The book ignores the fact that Porter’s work was shown to be lacking at least as far back as 2006.
  3. The simple analysis and lack of research into what is driving the shift results in trite recommendations.

The usual response to these points is are along the lines of:

  • ‘But everyone is using Porter still’ – which is an observations and not an argument.
  • They point out that the book is from a professor at Columbia Business School and published by HBR – which is just appealing to credentials.
  • ‘But the book is based on lots of analysis’ – which it is, but the analysis is riddled with holes.

Let’s handle the second point first.

Do Porter’s theories still work?

Porter’s work on competitive strategy might be one of, if not the most, cited works by business academics. This doesn’t mean that it’s any good.

Academia is riddle with frameworks that either have little or no evidence behind them, or which have been proved to be irrelevant in the modern context. This hasn’t stopped them being used as the foundation for new work.

Abraham Maslow and his pyramid of needs, for example, has been shown to have no basis in fact1 – it’s just something Maslow made up one day – and yet it’s taught in every b-school in the world. There’s similar problems with business value, technology adoption, and a whole range of topics.

Just because everyone uses Porter’s five forces doesn’t mean it works or has any basis in fact.

As Matthew Stewart pointed out in 2009 in his book The Management Myth,2 the idea of being able to locate and explot a sustainable competitive advantage was ‘lacking any foundation in fact or logic’. It’s rent seeking of the worst form. There’s an earlier article by the same author from 2006 that points out many of the same flaws.3 Even earlier in 2000 Pankaj Ghemawat conducted a survey of the history of business strategy which found that ‘In the case of the five forces, a survey of empirical literature in the late 1980s—more than a decade after Porter first developed his framework—revealed that only a few points were strongly supported by the empirical literature generated by the IO [industrial organisation] field.’4 The report he drew this from (and which I don’t have in my hands yet) is from 1989.5

The market even rejected Porter’s theories conclusively in November 2012 when Monitor Group, the firm that Porter founded to consult around his theories, filed for bankruptcy. As Steve Denning over at Forbes commented:6

Monitor wasn’t killed by any of the five forces of competitive rivalry. Ultimately what killed Monitor was the fact that its customers were no longer willing to buy what Monitor was selling. Monitor was crushed by the single dominant force in today’s marketplace: the customer.

It was Drucker who pointed out that the whole point of a company is to create a customer, not to try and squat in some magical place that would allow a firm to extract rents without any effort. Porter appears to have ignored this.

So no one wanted to buy the sustainable competitive advantage snake oil from Monitor, nor was Monitor able to apply the theory to its own situation and save itself. The idea has no basis in fact, the market rejected it, and it doesn’t work. And all this happened well before The End of Competitive Advantage was written or published.

Let’s set aside the idea that Prof McGrath’s book is the first time that Porter’s theories have been shown to lack potency: clearly she’s a decade or so late to that party.

The analysis

The second problem I have with the book is the poor quality of the analysis. Generally, the approach used by The End of Competitive Advantage is of the same level as Good to Great, which is another business bible that typically can’t be questioned but is riddled with holes. A lot of data might have been used, but the process is clearly deeply flawed.

The End of Competitive Advantage is built on a set of ‘growth outlier’ companies which out-performed the market. As is stated in the book:

In 2010, my research team tracked down every publicly traded company on any global exchange with a market capitalization of over $1 billion US dollars as of the end of 2009 (4,793 firms). Then we examined how many of these firms had been able to grow revenue or net income by at least 5 percent every year for the preceding five years (in other words, from 2004 to 2009).

These firms were then compared with their top three competitors and then with each other to identify what made them different. (Comparing a firm with its top three competitors is not the same as controlling for natural industry or geography growth, but we’ll let that one slide. At least there was some attempt to normalise the results. We can also set aside the question of why a five year period was used, even though it seems completely arbitrary.) The rest of the book presents what was learned, and provides the reader with some advice and a simple framework that you too can use to copy these growth outliers’ success. (This is why some reviews think that the book is an extended ad for consulting services, as the information presented is not much more than a teaser.)

As the book states:

The major conclusion was that this group of firms was pursuing strategies with a long-term perspective on where they wanted to go, but also with the recognition that whatever they were doing today wasn’t going to drive their future growth. Interestingly, they had identified and implemented ways of combining tremendous internal stability while motivating tremendous external agility, particularly in terms of business models.

The first issue we can call out with the analysis is a lack of disconfirming research. Consider, for example, if the CEOs of all the growth outliers wore red socks on Tuesdays. We might conclude that wearing red socks on a Tuesday will give us the edge we need. Humans have a natural confirmation bias so when you reach a conclusion you need to ask yourself ‘What would it take to prove this conclusion false?’ Can you find a significant number of firms where the CEO religiously wears red socks on Tuesday and which are not growth outliers? How do we know that the correlation they you’ve found isn’t just a happy accident, and that we’re reading a lot more into it than we should?

Next we have to consider survivorship bias. Someone has to win, but coming out on top does not imply that you were more skilful. There’s a lot of dumb luck in business; it’s not enough to be good at what you do, you need to be at the right place in the right time with the right product(s) and you still need a healthy does of luck. Did the growth outliers survive because they were good at what they do? Or is their success the result of happy accidents that took down their competition, or lucky coincidences that enabled them to leap ahead? Were they in the right places at the right time, moving into Asia when their competitors moved into South America, for example? Someone must survive, but there’s no rule that says that their skill was the only determinant of their survival.

Next we have the unknown unknowns. How do we know that the practices identified by Rita and her team are the right practices? Perhaps some of the outliers were more financially savvy and managed their cash flows, something which is hard at the best of times and even more challenging in the current turbulent environment, and which is inherently boring. Or perhaps they made a couple of astute (or just plain lucky) bets on which sectors to play in, nudging them past their competition. How do we know the survey or practices was complete? What was the framework used to identify these practices, and link them to changes in the environment. Correlations don’t cut it.

Ultimately, identifying a common set of practices for a set of companies that performed well over a given time period does little more than confirm that over the last time period these companies did well. That was already obvious.

We need to build a model that allows us to feed in long term market trends (increasing competitive intensity, decrease in ROA – at least in the US – blurring of sectors, etc.) and ask questions like, ‘How would these companies have performed three or more periods back in the past, and how might they perform in the future as the market evolves?’. If we’re looking for a change in the market then there should be an earlier time period where these practices were counterproductive. It’s this sort of approach that makes Thomas Piketty’s new book so interesting.7

If you’re going to write a book about what to do in the future than you need to do more than point out what worked in the past, even if it’s the recent past. The future, as they say, is a foreign country.

This is also the big mistake that Good to Great made: identify a group of profitable companies that have some shared characteristic, assume that what made them successful in the past will also make them successful in the future, and then call out the common elements from this set of companies. As Freakanomics, put it, a lot of the Good to Great companies went ‘From Good to Great … to Below Average’.8

The recommendations

Given all this, the book introduces the idea of ‘areas’ as the basis for competition, rather than industries. Its a nice idea as it allows us to pull more context into our analysis of the market: industries modulated by a few different dimensions, such as geography, demographic, etc.

The concept falls down, though, as it ignores the fact that industry definitions have become fluid. (What? Apple is a PC maker, not a phone maker? And what’s this touch and apps store stuff?) This  means that areas must also be a fluid concept, but the book does not look into the dynamics of how areas change. (I expect that this is left as an exercise for the reader, or they assume that you use Porter’s model to evaluate opportunity.)

The model for managing change across areas is a simple launch, ramp-up, sustain, ramp-down, disengage process. This doesn’t account for the pace in the current market. If your competitor can launch a new product in two to six weeks from a standing start (as many companies now can) then, while your carefully thought out launch process taking six months might seem modern, it’s largely irrelevant. There is no insight in the recommendations on what the change in market pace means other than ‘the end of competitive advantage’.

The recommendations all but ignore the shift from stocks to flows,9 which has huge implications for how we think about, organise, govern and manage our business. There is, however, a brief mention to the idea of consuming services rather than building assets, and the book even name-checks Odesk. However, it doesn’t look into the implications that spring out of this. The coverage is only a few spare paragraphs and you’re left wondering if the author doesn’t really know what to make of the topic.

The rest of the book which follows is a fairly straightforward process of working through the stages of the model and providing a few points of sage sounding advice for each stage (‘Rotate you team through departments so that they don’t get comfortable’ type of thing). You’ll either nod and say yes to each of these (the Barnam effect10 in action) or go ‘meh’. Your mileage might vary.

My review

So, as you can see, my opinion is based on the following pillars:

  • The main thesis that ‘sustainable competitive advantage is over’ is old news.
  • The analysis is suspect, at least, and doesn’t prove the thesis.
  • The model and recommendations provided hold little value.

Calling it ‘crap‘ might be going a bit far, but for me the book was a waste of money. Use the money to buy a coffee for a friend that you haven’t spoken to in a while; you’ll learn a lot more.

While the content might come from a major b-school and has been written up in respected journals, that doesn’t change the fact that we live on the internet now and we need proof that we can see. As Jay Rosen pointed out the other day,11 appealing to credentials doesn’t work in this day and age.

  1. William Kremer & Claudia Hammond (31 August 2013), Abraham Maslow and the pyramid that beguiled business, BBC World Service. ↩︎
  2. Matthew Stewart (2009), The Management Myth: Management Consulting Past, Present & Largely Bogus, W. W. Norton & Company. ↩︎
  3. Matthew Stewart (June 2006), The Management Myth, The Atlantic. ↩︎
  4. Pankaj Ghemawat (April 2000), Competition and Business Strategy in Historical Perspective, HBS Comp. & Strategy Working Paper No. 798010 ↩︎
  5. Richard Schmalensee, ‘Inter-Industry Studies of Structure and Performance’, in Richard Schmalensee and R. D. Willig, eds., Handbook of Industrial Organization, vol. 2 (Amsterdam: North-Holland, 1989). ↩︎
  6. Steve Denning (20 November 2012), What Killed Michael Porter’s Monitor Group? The One Force That Really Matters, Forbes. ↩︎
  7. Thomas Piketty (10 March 2014), Capital in the Twenty-First Century, Belknap Press. ↩︎
  8. Steven D. Levitt (28 July 2008), From Good to Great … to Below Average, Freakonomics. ↩︎
  9. Peter Evans-Greenwood (20 Feburary 2014), Setting Aside the Burdens of the Past, PEG. ↩︎
  10. From wikipedia: The Forer effect (also called the Barnum effect after P. T. Barnum’s observation that ‘we’ve got something for everyone’) is the observation that individuals will give high accuracy ratings to descriptions of their personality that supposedly are tailored specifically for them, but are in fact vague and general enough to apply to a wide range of people. This effect can provide a partial explanation for the widespread acceptance of some beliefs and practices, such as astrology, fortune telling, graphology, and some types of personality test. ↩︎
  11. Jay Rosen (March 2014), “I want it to be 25 years ago!” Newsweek’s blown cover story on bitcoin, PressThink. ↩︎
Tags