As the adage attributed to management guru Peter Drucker goes, “If you can’t measure it, you can’t improve it.” Well, you might be able to improve it by chance or by heroic effort, but not in a sustainable way. In business we tend to measure and improve safety, quality, delivery, cash, cost and productivity, as well as morale in the form of employee satisfaction or employee engagement. Productivity seems to be the performance metric that we discuss most casually, yet are often confused by in terms of its meaning. Perhaps this is because while safety, quality, on-time delivery, lead-time, cash / capital levels, and cost are what we might call simple quantitative metrics, and morale metrics are subjective qualitative metrics, productivity is a ratio of two different things; value outputs and time inputs.
Productivity Declining – Or Not?
There was a very interesting article last week in the Wall Street Journal titled Silicon Valley Doesn’t Believe U.S. Productivity Is Down which questioned whether the recent downward trend in U.S. productivity, after decades of steady growth at around 2% per year, was real or just a measurement problem. Economic minds from Silicon Valley attempt to make the case that high productivity contributions from technological innovations are not being accounted for based on how the government measures productivity in the U.S. Why productivity improvement appears to be slowing down recently is an interesting question. At best, the Silicon Valley arguments come across as, “That feels wrong, so it must not be true.”
The productivity discussed in the article is the dollar-measured output of goods and services of the U.S. economy divided by the number of hours worked by U.S. workers. The measurement discussed in the article is limited to the business sector, about 75% of the U.S. GDP and excludes government, non-profit work, private household employees and owner-occupied rentals. In simple terms, productivity is a measure of value we deliver from a given effort. There is an inherent problem in that some goods and services are free, are priced incorrectly, or change in price due to inflation and deflation. Economists point out that these things tend not to matter because we are comparing the past performance of the U.S. economy to the current performance, so these imperfections still give us a consistent measurement of growth and trends.
The perspective from Silicon Valley is that much of the economic value created there is not being captured properly in the U.S. government’s productivity metric:
One measurement problem is that a lot of what originates here is free or nearly free. Take, for example, a recent walk Mr. Varian arranged with friends. To find each other in the sprawling park nearby, he and his pals used an app that tracked their location, allowing them to meet up quickly. The same tool can track the movement of workers in a warehouse, office or shopping mall.
“Obviously that’s a productivity enhancement,” Mr. Varian says. “But I doubt that gets measured anywhere.”
No, that is not obviously a productivity improvement.
It seems that Mr. Varian is confusing quality of life improvements, or convenience, with economic productivity improvements. In life, working people broadly divide their time between sleep, leisure, work and what we might call maintenance – a range of activities including grooming, cleaning, and other needed biological or social functions that might not fall into the leisure or work categories. Productivity only measures dollar value output and time input for work. Productivity does not care that we have apps that let us find our friends in a park more quickly, if this is a leisure activity with no dollar value output. Simply being able to find workers in a workplace more quickly, or even identifying movement patterns that were unproductive, does not increase productivity. Unless the process of tracking people with the app allows a business to reduce input per unit of output, it is a convenience but not improved productivity.
When we look at the flow of value from process to process, from customer request to fulfillment of that request, this is what we call “end-to-end thinking” in the lean management vernacular. This is far from obvious to most people. When we make pockets of improvement, such as the one described by Mr. Varian in the article, they result in it is a local (which is to say false) efficiency, at best. At worst, they reduce the productivity of the entire system by speeding up one process to create excess waiting or excess stock because the next process in the chain has not been yet been similarly improved. Technology may reduce the time it takes for us to accomplish certain tasks, but how is that time we saved reallocated? Time is unlike money, where “a penny saved is a penny earned”. Time saved is not time saved, if we do not use the time for something else useful. We can’t bank time. We consume it intentionally or by inaction.
Jumping to Conclusions in the Valley
Exciting new technologies bring with them optimism. This is generally a good thing, unless there are no rational behavioral economics present to warn us against our cognitive biases that arise from an atmosphere of enthusiasm. A Stanford economist is quoted in the article,
“You can’t be in the Valley without thinking we’re in the middle of a productivity explosion,” Mr. Bloom says. “And when they do discuss it, everyone jumps to Hal’s conclusion here.”
You also can’t live your life on an active volcano without thinking that your world is heating up, but that is a poor way to judge the average air temperature of the rest of the country. It’s sad that Mr. Bloom doesn’t see a problem that smart people in Silicon Valley, highly intelligent engineer and scientist-types, are allowing themselves to “jump to conclusions” rather than arrive at a conclusion through a logical chain of causation. While it may be true that Silicon Valley businesses are far more productive in terms of dollar output per worker hour, the U.S. government measures showing the country’s productivity declining can’t simply be wished away with technological optimism. There are phenomena there begging to be studied and understood.
It’s quite possible that people in Silicon Valley blur the lines between work, play, grooming and sleep, and technology-fueled time savings in any one of these areas naturally leads to wiser use of time and greater productive output. I can imagine that time saved in ordering food or hunting for a taxi is spent to talk about creative solutions to a product development problem over dinner. As more of the economy shifts towards knowledge-based creative service economy, it may become true that time saving apps generate economic productivity by enabling more of us to capture and reuse pockets of time savings for valuable economic output. For those of us who rely on physical capital such as farmland, vehicles, machine tools or retail space to create value in our work, pockets of productivity improvement still offer little overall benefit.
Taiichi Ohno’s Admonition
Lean management guru Taiichi Ohno warned against the illusion of productivity. When he was leading efforts to improve productivity at Toyota, he made a clear distinction between the false productivity and true productivity. For example, a group of 10 people able to produce 100 units in one day become more productive and are able to produce 110 units per day. Ohno argued that if the customer demand was 100 units, it was a bad thing to produce 110. Instead, 9 people should work the day to produce only 100.
Furthermore, Ohno said that we should not produce these 100 unless they were actually sold. Ohno warned against overproduction, or output which is not consumed. This leads to increased costs from inventory management, obsolescence, price reductions due to oversupply and the obscuring of the true status of the work process. Flow is what connects a local optimization to an output and thus the ability for the pocket improvement to be realized as productivity gains. This insight was at the basis of his design for the Toyota Production System.
Armies of traveling salesmen who catch Uber cars to save 20 minutes waiting in the airport the taxi queue, only to arrive 20 minutes earlier to their sales calls, gain no improvement in productivity. There is nothing wrong with arriving early, and it is even preferable. This trading of one variety of wasted time (waiting) for an identical one in another place in the process is a typical challenge of achieving true end-to-end improvement. Ohno would have looked at the whole process, from acquiring the prospect to the sale and collection of cash, and asked where improvements were needed to improve productivity.
Let’s Solve Whole Problems
Google founder and CEO Larry Page said one of his 5 keys to startup success has been to “solve whole problems”. The best services, technological or otherwise, solve whole problems for customers. These are based on an understanding of customers and what type of service and experiences they want. We could not imagine a Google search product that delivered the search result we wanted, but didn’t allow us to click on the link, or an Amazon store that allowed us to order the product but forced us to drive to a brick and mortar store to pick it up, or a Microsoft Office product that didn’t allow us to print out documents. These would be far less than whole solutions.
Even the best technology solutions address only a part of the value delivery process, which extends from customer request to fulfillment of that request. As it is measured today, we are not more productive if we quickly find what we want on Google but are not able to use that information to solve a real life problem. We are not more productive if we are not able to use the printed output from Microsoft Word to generate the desired document for a customer. We are not more productive when Amazon saves us time by delivering the home improvement tool, if aren’t prepared to allocate the time saved going to the hardware store to generating other valued outputs. The challenge with improving true productivity is that almost none of us work end-to-end. Most of us still work in silos, able to see only a small part of the overall process that we are part of, and able to influence less of it than we would like. This will not change any time soon. What can change is that we choose to work a team, to take the customer’s viewpoint and to look for whole problems to solve.
Don’t worry Silicon Valley, declining U.S. productivity still believes in you.