Optimization. Experimentation. The incorporation of data into decisions will lead managers to a fundamental decision. Do they value optimization or experimentation? Every company, every country will face the same value decision. In many ways this is an old question. Should a company reinvest profits or distribute dividends? Should a country reinvest its resources into infrastructure and the private sector or redistribute resources?
Amartya Sen tells us this is a false dichotomy. He teaches us to invest in human capital. This form of reinvestment serves provides economic redistribution while producing economic gains. This is the Google or Facebook model. Their commitment to experimentation has become a competitive advantage. They use experimentation to learn new ways to optimize their revenue streams. But this is a naïve view of business. It’s easy to make rosy projections from the confines of the classroom. The rest of us are working in an office in the morning and sweating in a warehouse in the afternoon.
Net profits are the measure of success in a mature industry. The gains from experimentation are slim in many industries. But efficiency always has value. The new data revolution will lead these companies to pursue optimization. But will it produce experimentation?
The tech giants like Google and Facebook balance between experimentation and optimization because their sample sizes are vast. This contrasts with the typical small business. In the book Trust, Francis Fukuyama described the Keiretsu system in Japan. This book is written from an academic point of view. Anyone working in the American industrial sector recognizes the interlocking business relationships. Most small job shops survive from just a few customers. The presence of a major assembly plant such as Caterpillar, Toyota or General Electric will lead to multiple suppliers and contractors within the community.
A small medical manufacturer was a customer of my company for over ten years. They survived the Great Recession but folded a few years later, as the larger economy continued to improve, because they lost a single customer. It turned out 95% of their revenue was tied to this customer. Their entire business model was built on satisfying this single customer.
As the tech industry matures, the FAANG companies will transition from stars into cash cows. It is likely experimentation will produce smaller gains. They will likely focus on the optimization of their business model. Net profits will demand a reduction in any expenses that do not contribute to immediate revenues. This transformation changes the calculus of a business model and makes it vulnerable to product substitution or innovation from outside companies. The emergence of Tesla is possible because it has redefined the design of automobiles into a growth industry. Moreover, he has redefined the terms of the product into technology rather than machinery.
Governments face a similar dynamic. Sen makes it seem as though income redistribution produces economic growth. This is an oversimplification. Countries will face diminishing returns as their spending in education or healthcare increases. Developing nations will see dramatic improvements initially. But a developed nation with high standards of education and healthcare is unlikely to see the same gains. This does not mean the gains are nonexistent, but they will not be to the same degree.
The United States might be an exception because education and healthcare are inconsistent between social class and communities. But these investments must be targeted to the communities who need them to gain its maximum effect. And yet no government is immune from the basic decision data will provide. Nobody will accept experimentation. Everyone wants to optimize public resources.
jmk, carmel, indiana, email@example.com