Lean thinking and practice have the potential to improve outcomes in healthcare, software development teams, government services, business performance and even individual growth and well-being. For something that is so powerful and effective, Lean has many ways in which it can fail. This is because contrary to popular belief, Lean is complex, not simple. This is not to say that it is complicated or difficult to understand. It is complex in the sense that there are many interacting parts to it. Lean is both a system of systems and the application of system thinking to how we live and work. Naturally, Lean fails when we treat it as a micro solution or technology to be bought and implemented rather than as a system to be tended through its seasons, like a garden.
A recent conversation with a senior leader of an organization practicing Lean reminded me of the productivity paradox of information technology. His firm is dutifully doing kaizen events, implementing Lean tools, practicing and auditing 5S, beginning to hold tier meetings to expose and solve problems. Things look good, people feel good and there is a general sense that they are going in the right direction. But they are not seeing results on the bottom line. In fact, this senior leader is hearing voices that some productivity measurements seem to be trending down after launching Lean. While they are not yet saying this is due to Lean, they are scratching their head at this paradox. A similar productivity paradox was observed by economist Robert Solow in the first decades of information technology investment, the 1970s through the early 1990s. As organizations made IT investments, worker productivity appeared to go down instead of up. To rephrase Solow, the senior leader’s concern was essentially, “We see Lean everywhere except in the productivity statistics.”
Historically the return on investment in terms of productivity from mechanization and automation of the agricultural and manufacturing sectors was in the 3% – 4% range. People expected similar returns from IT, but it was around 1% from the 1970s to the early 1990s. Quite a few theories attempted to explain the paradox of “improving technology but stagnating productivity”, including problems with how we measure, a time lag before gains were visible, that office people found other things to do with time saved, and the changing nature of what is “information technology” as compared to the relative stability of what we mean by mechanization. We did see higher productivity gains after the 1990s, and this is often attributed at least in part to IT, but I would not declare the Solow paradox a settled matter.
What does this mean for Lean practitioners, and how we can avoid the so-called productivity paradox?
First, it’s important to be clear-eyed about the nature of what we are measuring. Productivity is at best an aggregate and limited measure of progress. It is total output divided by total input. We need to consider safety, quality, on-time, agility and how engaged people are in the enterprise. Cost is equally tricky to pin down and link directly to IT or Lean investments on a cause-and-effect basis, because the many inputs and outputs within the system blur the picture. When we make it clear what specific problem we are solving with IT or Lean, we have a better chance of tracking cause-and-effect.
Second, this senior leader observed that his people seemed to be trying to improve everything and anything, even if by a little bit. Was this time allowance for continuous improvement hurting productivity? This is another common pitfall of Lean. On the one hand, “everyone everyday improving everything” is a good long-term vision statement. On the other hand, in the near-term it is a good way to bankrupt your Lean effort. Yes, we should strive to improve everything eventually, but in the beginning the best way to see results is to only improve what matters. Just as it doesn’t make sense to “go paperless” and put in fancy IT systems for every process regardless of the cost to do so and the muddiness of expected benefit, we need to be careful where we invest our process improvement energy. Organizations that are already profitable may benefit from focusing less on results from Lean at first, letting people experiment, get them engaged in improving everything and anything even just a little, and bringing structure and focus later after building acceptance to improvement. The loss of perceived freedom can result in a backlash, if the Lean engagement is too loose and lacking top-down direction at first. This is a tricky topic requiring forethought and good judgment.
Third, how do we know what matters? Ask customers. The start of any Lean strategy should be identifying value streams and listening to what the customers of these streams want from them. If we want to see results, it is advisable to improve things end-to-end horizontally across business processes. Upgrading from manual typewriters to word processors helps nobody if the orders that we type up faster, better and cheaper, just sit in a warehouse clerk inbox waiting to be read and picked. Seeing improvements on the bottom line, whether productivity gains from IT or profit and cash improvements from Lean, requires that we link up improvements side-by-side from customer request to fulfillment, one value stream at a time. But this requires customer-first thinking, cross-functional cooperation, and an openness to try new things. These are not a given. Developing this mindset may require the aforementioned period of “improve anything and everything”.
There is only a productivity paradox when we expect linear and mechanical cause-and-effect relationships between investment in systems and improvements in their performance. With a basic understanding of the differences between a machine and a system, we can see that such an expectation is not rational. It helped my senior leader friend to hear of Solow’s paradox, the parallels between how we adopt IT and Lean, and the lag in measurable results from positive changes. He realized that the challenge organizations have in seeing results from Lean is not necessarily a feature of Lean itself. It is in the nature of systems, cause-and-effect chains and how we measure productivity.