Even a bad strategy can do good to business

Deviations from official strategy sometimes result in huge gains for companies just as Apple found out with its Graphing Calculator software and Google with Gmail.

When an assignment to create the first version of Apple’s Graphing Calculator software was cancelled in 1993, freelance software developers Ron Avitzur and Greg Robbins paid no heed. In an act of innovation-as-rebellion that has become legendary, they used their Apple ID badges to gain unauthorised access to the company’s campus, working into the wee hours for six unpaid months until the project was finished. Ten years after its completion, the Graphing Calculator software had shipped with an estimated 20 million machines.

This is a compelling example of what organisational researchers call “bottom-up exploration” — employee deviations from official strategy that sometimes result in huge gains for companies. Apple isn’t the only Silicon Valley firm to have benefited from letting staff follow their muse: Google famously allows its employees to spend 20 per cent of their time on company-related personal projects, a policy that led to Google News, Adsense and Gmail.

Knowing as we do these benefits of deviations from strategy, however, as well as the reality that the strategies coming from the C-suite are seldom perfect, is it sensible for managers to place such a heavy emphasis on implementing them effectively? In a recent working paper, Explaining the Implementation Imperative: Why Effective Implementation May be Useful Even with Bad Strategy,Eucman Lee — a Ph.D candidate at London Business School — and I develop a theory that aggressively pursuing effective implementation may in fact be very sensible. By “effectiveness” at strategy implementation, we mean the extent to which an organisation’s actions correspond to its strategic intentions. Thus a company that seeks to pursue a low-cost strategy can be said to have successfully implemented the strategy if its costs indeed fall relative to its rivals.

Whether this leads to high profits or not depends on the appropriateness of the low-cost strategy in that particular industry.

The fundamental feature of strategy implementation that we focused on in our research is the separation between beliefs and actions. In the typical company the people who come up with strategies and refine them are not those who implement them. In an attempt to study the consequences of this separation carefully, we built what is known as an “agent-based model,” basically a computer programme that replicates the logic of interaction between individuals in a way that allows us to project what is likely to occur in the course of many such interactions in a wide variety of circumstances.

Improve the implementation

Our model involved a manager and a subordinate, programmed to try to look for the biggest possible profit by choosing from a range of options through trial and error, akin to a gambler facing a slot machine with several arms. Each period the manager would pick a strategy and “tell” the subordinate what to do, and the subordinate would implement the strategy as he understood it. There would be a performance outcome, and then the manager would modify his beliefs about the value of the strategy based on the performance observed.

We ran the model through numerous periods, building in different types of features corresponding to the real world such as communication errors and top-down and bottom-up exploration of ideas.

Across a range of conditions, we found that it was generally a good idea to improve the implementation effectiveness of the subordinate, even when the strategy the manager chose was not a good one to begin with.

As we picked open the model to see what was going on, we discovered two main reasons for this phenomenon.

First, bad implementation makes it difficult for companies to learn from either failure or success. When a strategy produces undesirable outcomes, how are leaders to know whether the problem lies in the strategy itself or in all the deviations that crop up in the absence of effective implementation? If the outcome is good, how do we know if it was indeed because of the strategy? This could lead a CEO into unfortunate decision-making based on a confounded impression of the outcome.

Second, the organisation as a whole does indeed benefit from learning better strategies through some deviations from current strategy. Beyond a certain point these aberrations hurt, however, because they don’t allow the organisation to extract the value of the good strategies uncovered. Any communication gap between managers and employees automatically will foster some amount of divergence, and attempts by senior managers to look for new strategies also generate deviations in the course of time. On top of these, deviations resulting from imperfect implementation tip the level of deviation into the harmful zone.

Measuring the process is vital

Our results also suggest that companies not only should continue to invest in improving their strategic implementation but also should focus on sharpening their measurement of implementation effectiveness. Indeed, a manager who looks, listens and accurately interprets implementation effectiveness can be a greater asset than a silver-tongued boardroom orator who knows how to communicate the strategy effectively.

Why is that? The well-communicated strategy may not be the best anyway, and deviations arising from misunderstanding it can be benign. However, an unobservant manager may contribute biases, false realities about what actions actually were driving current performance. Indeed, eagle-eyed managers who can measure implementation effectiveness are the most likely to help companies capitalise on innovations originating from bottom-up exploration. The potential breakthroughs that occasionally come about when employees wittingly or unwittingly deviate from company strategy are unlikely to be replicated, let alone propagated as best practice, without managerial intervention.

Managers instinctively know that good implementation is important, but too often they think of it as being only as good as the strategy it serves. In fact it should be considered an adaptation mechanism in and of itself, not merely a way to bring subordinates into line but also an essential tool to help find and fix flaws within the current strategy — and find better ones.