What have we learned so far? In the first video, we have seen how technology evolves over time. We have discussed the three fundamental laws and how they translate into falling transaction cost. Then, we saw the impact this has on the broad business architecture of some industries. We specifically talked about the phenomena Philip Evans calls the deconstruction of value chains, something that we will come back to later. For now, let's go one layer deeper and ask the question, how does technology impact the performance of specific companies, if at all? We will start with the famous paradox, usually called the Solow Computer Paradox, for Robert Solow. He's a Nobel Prize laureate in economics, and he has been quoted saying, you can see the computer age everywhere but in the productivity statistics. This is a very powerful claim. Very controversial, but it is not completely unjustified. In fact, when you look at the productivity growth in the last decades, you see that it has not kept pace with the growth in information technology spending. What I mean by information technology spending here is the total corporate investments in hardware, software, data centers, networks and the related human resources. This amount is nearly $6 trillion per year. And it grew by a factor of almost 20 between 1980 and 2015. In the same time frame, global GDP barely tripled. This is exactly what the paradox refers to. Inferring that technology investments didn't help us create more economic value. For any company or any rational economic agent, this would mean they should stop such investments. And I guess you see why the initial claim is a paradox. Because both intuition and practice are going against its natural conclusion. So far, I have not yet met a CEO whose turnaround plan starts by stopping all IT spending and going back to the 1920s process manual. Now, before going any further in the course, I want to take time to establish with you that, for companies, technology is a massive value creation lever. And I will split this discussion into two parts, first theoretical and then empirical. The theoretical discussion. I will go in there one by one through the most important attempts to solve Solow's paradox. First attempt, maybe technology had a beneficial impact, just not on economic productivity, not on GDP, which is a very narrow measure. Second attempt, maybe it had a positive impact on GDP, but it will only show after a long time lag. Third attempt, maybe it had a positive impact on GDP and in the short term, but it was neutralized by some other business phenomenon. Let's start with the first measurement argument. Most of us would agree that technology has obvious social and emotional benefits. My mother could see and talk to my daughter anytime, despite the two of them being 11,000 kilometers apart. But I would argue that work productivity benefits are equally obvious, at least at the individual level. Just look at the time it took to design mechanical components a few decades ago. You needed to solve complex differential equations by hand with a series of approximations to know how the component will behave under this or that kind of pressure. Now, you run this through a numerical model in literally a few seconds. This has led to more efficiency, stronger risk management, and ultimately it led to better and more innovative, more value-adding products. So although GDP is a narrow measure, it is fair to expect that technology impact would at some point trickle down there. Moving to the second argument. If we accept that technology had a productivity impact, maybe it doesn't show immediately because it simply takes time. This is a reasonable expectation. In every massive change, we should expect that only a small fringe of a company or of society would be early adopters. For those who have run large scale transformation programs before, you know that it takes time to get the majority on board. To convince and train employees to use new technology, or a new process. And even after that, it takes time for those employees to be fully productive with the new tools. This might also explain the delayed bump in productivity growth we saw in the late 1990s. And it would mean that we are already in times where technology impact is visible. Now, the third argument is a bit more subtle. It goes something like this. If we accept that technology had a positive productivity impact, that it is visible today, maybe the reason we didn't see it is that there are other factors impacting productivity, but this time negatively. In such a way that the total measured impact is neutral. If you are thinking, what factors might have such an impact? There is one that comes naturally to mind, complexity. As an illustration, today, on average, companies set themselves six times as many performance requirements as they did in 1955. Back then, CEOs committed to between four and seven performance imperatives. Today it's between 25 and 40. And actually, many of those requirements can be contradicting. When it was enough for a car to be safe in the 50s, today it needs to be also innovative, cheap, have a good brand name, easy to maintain, and have tons of functionalities. So the increased complexity in business environments might have neutralized the positive impact of technology on productivity measures. To summarize the theoretical discussion, the Solow computer paradox doesn't refute the positive impact of technology. It points out that part of this impact is non-productivity related, that it can be delayed in time, and that it is covered by the negative impact of increased complexity in today's companies. Before we move into the empirical discussion, I would like to share a story highlighting its importance. I was having a discussion with the management team of an Asian central bank, precisely on the impact of technology. After debating the three arguments and spending some time on how complexity is really changing the regulation of markets, the deputy governor asked a thought-provoking question. Isn't technology itself what creates the complexity it is trying to counterbalance? Well, although I don't completely agree, I can see how part of the complexity is enabled by the fact that consumers are now able to compare competitors across markets. That employees are expected to master more and more tools. And that managers are juggling new channels of communication and coordination. This is where the empirical side is key. We won't reverse the technology trend, even if we establish how much complexity it creates, simply because it is compensated by the value creation potential. What value does this digital technology create for a business, empirically speaking? There are multiple studies researching this very topic. And the answers, of course, vary by industry and depend on the starting position of specific companies. In a BCG perspective published last year, we saw a positive correlation between technology investments and gross margins. Top performing companies tend to have higher technology intensity index compared to the industry average. These results go in line with another Harvard Business Review study focused on financial services industry, where digital leaders outperform digital laggards in customer loyalty, and ultimately in revenue growth. This holds true for small and medium enterprises as well. A 2013 BCG study found that technology adoption leaders outperformed laggards by 13 percentage points in yearly revenue growth. This was in developed markets, in emerging markets that number is 15 percentage points per year. This can help many of you build more than solid business cases for your next technology investment plan. Let me now recap the most important messages from this video. Although we don't necessarily see it in productivity numbers, digital technology does have a visible impact on business performance. Some of this impact can be neutralized by the increased complexity generated by competition, customer demand, or regulation. Lastly, in practice, making the right technology investments translates into higher profitability and higher revenue growth. Double digit higher.
