The fall of Kodak is a tale of missed opportunities, denial and ultimately crushing failure.
Now a case study in corporate short-sightedness, Kodak was the camera king. When taking pictures involved a roll of film and sending it off to be processed, Kodak was number one. In 1990 its annual sales were $19 billion and it employed 145,000 people worldwide.
It’s yellow and red logo was omnipresent. Just as Google has become a verb for searching on the internet today, so the phrase “Kodak moment” was part of everyday parlance to describe something so worthy it should be immortalised by a photo.
But the American firm was sent spinning into bankruptcy in 2012 because of the digital camera, negating the need for film and Kodak’s business model. This is despite Kodak inventing the digital camera in 1975 and producing the first mega-pixel version in 1986.
Kodak’s demise has become the archetypal salutary lesson for all business executives of the ‘competency trap’. The trap is that a company becomes so good at one way of doing business it ignores new technology or methods which might be more difficult in the short-term, but will eventually produce better results. Its short-sightedness ignores the long-term benefits of pursuing the new method or technology.
For Kodak it was the digital camera. It did produce a digital camera, but it made a loss on every one it sold and so continued to invest most of its resources in film photography. While it did this rivals were busily perfecting the digital camera. And when consumers finally switched, the decline in sales of film was rapid, leaving Kodak with a vast manufacturing base in old, redundant technology. Its competency was no longer in demand.
Much has been written on why Kodak could not escape the competency trap, with companies desperate to learn the lessons. Technological disruption, thanks to the digital revolution, has put most firms on high alert to avoid a similar fate as they look to maintain the right balance between exploitation of what they are good at and exploration of the technology that can change their industry.
But we have found that the competency trap is not really a trap at all. By building models using algorithms we have found that competency is perfectly logical. At the time when a decision needs to be taken, it is the right path to take. Companies should fall into the trap - it really is the rational choice.
Ignoring the long-term benefits of the new technology or method is not myopic, it is, at the time, the smart choice. Those advocating the competency trap are often doing it from the privilege of looking back on the past choices companies’ made, with all the abundant information that would eventually become apparent. They are suffering another well-known bias – the hindsight bias.
To test the competency trap we used learning algorithms that are not myopic and do consider the long term, plus, they take on substantial exploration. We did thousands of simulations with the algorithmic models to test various scenarios. Essentially, there were four routes from the outset: 1) it starts bad but improves; 2) things start well but performance declines; 3) it starts well and is always good; or 4) it begins badly and stays bad.
Even when we tweaked the algorithm so it was rational and knew that using the alternative route would initially be poor, but would improve, it still chose to fall into the competency trap. This rules suggests short-sightedness is not the reason why businesses fall into the competency trap and indicates it is impossible to ‘debias’ people against it.
How do you avoid the competency trap?
Indeed, there is plenty of research showing that managers do think in the medium and long-term. Kodak, after all, was always planning ahead. In 1981 when Sony produced the first electronic camera Kodak researched extensively the adoption rate and likely path of the digital camera. Kodak concluded it had 10 years to prepare before digital would take over film. This proved pretty accurate, but still Kodak was caught out.
The reason companies and people fall into the competency trap is because of logic. When looking at two alternatives, with one producing good results and the other bad, there is only one choice. There is no way to know then that the more inefficient, costly, harder and less profitable route will finally come good. It could be many months of losses, so sticking with the profitable and high-performing choice is perfectly rational.
After all what looks bad initially, could always be bad, so to keep on trying it is totally irrational, it would make sense to quit, but then the business won’t find out if it would have improved. This the asymmetry in the information of the two options.
And yet even if you know about the competency trap, how long can you keep ploughing resources into alternatives that are performing poorly - they may always remain bad options.
A lot of the management research that find firms suffering the competency trap is historical and can’t see at the time what the evidence was for making decisions. Kodak may not have invested more in digital cameras because there wasn’t enough evidence. What about other firms around that time that invested too much in digital cameras early on and perhaps went to the wall trying to make this expensive new technology profitable?
Looking historically at the route of the successful companies does not always explain why they succeeded, luck may have played a part. But when researchers look back they write as if it was inevitable. Often these are exceptional cases and Kodak may be an exceptional disaster.
Indeed, our models show the competency trap is expected and the outcome of rational investing. The only way to counter it is to become irrational. This is something I conjectured about in a paper with the man who discovered the competency trap, James March, of Stanford University, who died in 2018 at the age of 90.
One way of avoiding the competency trap, we argued, was to indulge in longer experimentation with alternatives that seem poor. Most would, however, stay poor and so it would be a very costly strategy.
Investing in substantial exploration, though, is seemingly the only way to avoid the bias, but as our algorithmic models show, it is not the optimal choice. Managers need to overestimate the alternatives to persist with them long enough. Indeed, those with an irrational belief in the value of persistence will be more likely to discover when persistence does in fact pay off. Entrepreneurs are often found to have such irrational belief in their product, despite many failures and setbacks.
It is only this kind of irrational zeal that will avoid the competency trap, but at such cost that, as our models show, this is one trap no firm should try to avoid.
Denrell, J. and J. G. March (2001), ‘Adaptation as information restriction: the hot stove effect,’ Organization Science, 12(5), 523-538.
Levinthal, D. A. and J. G. March (1993), ‘The myopia of learning,’ Strategic Management Journal, 14(S2), 95-112.
Denrell, J. and G. Le Mens, HERE (2019), 'Revisiting the competency trap', Industrial and Corporate Change, 29(1), 183-205.
For more articles on Strategy and Behavioural Science sign up to Core Insights here.