How big will companies be in the 21st century?

There are two incompatible narratives about how startup techniques will eventually be incorporated into the broader economy. One is that the startup period is basically over, and everything in the future will be done by the current crop of tech giants, a return to the concentrated corporatism of the 1950s. The other is that the internet has permanently altered communication costs, and small companies will always have an advantage over larger ones.

It’s important to note that we’re in a transitional period of some kind. Interest rates are still very low, and were zero not long ago — meaning the cost of borrowing capital is very low, because it isn’t being spent. Corporations have been hoarding cash because they don’t know how to spend it in a productive way. There are several theories of why this might be.

Most startup people would say that Google can’t spend that cash itself because big tech companies are not good at building original in-house software, which is perhaps why they’ve been buying startups and adopting the techniques of startups. It’s even tougher for older companies, which may be able to survive by acquiring the right tech, but may end up going under and being eclipsed by new companies which are built ground-up for startups.

A convincing theory is the zombie firms hypothesis, which suggests that the existence of low-productivity firms propped up by cheap credit deters newcomers from expanding into their verticals. The knockon effects from these firms could be enough to explain why companies like Google have productivity and profits that are very high yet don’t expand. If this is true, then a side effect of QE, which was intended (and worked!) to prevent mass unemployment after 2008, might have been to slow down creative destruction.

A peek of where we might be headed is with Amazon. Amazon isn’t sticking to the book business — it’s aggressively expanding it’s logistics operation to every area of commerce. Lately, it hired a huge number of people — 100k in the last year, probably the same number this year. However, Amazon also uses contractors (most obviously things like UPS) for much of what they do, so perhaps the big corps of the 21st century will directly employ fewer people than those of the 20th.

Sure, Amazon is a software company, but it’s also becoming this logistics backend for small retailers everywhere. This reddit comment explains the strategy well:

To some extent, this is what Facebook and Google already do with advertising: they’re big platforms that empower niche businesses. There are still more these platforms can do, and probably more such platforms still to emerge.


So what does this all mean? Are we headed towards a mega-corp-world dystopia, or should we cheer on the high-res society?

If I’m correct, if we follow the leading indicators, it looks like within the next 10–15 years, even more non-software low-productivity zombie firms (everything from food delivery to CPG to selling homes) will go out of business, allowing tech startups to fill their place. Startups that figure out a repeatable model for growth will be acquired by big software platform companies (and some of them may still yet become one of them).

These software platform companies will eventually use their profits and newly acquired know-how to expand, which will mean hiring more employees or contracting firms to do more real world stuff. They will be the big nodes that the economy will orbit around. There will be plenty of room for small, niche businesses that do one small thing really well, probably depending on local know-how.

Interest rates will rise and it will be hard to raise VC for a startup again. Big companies will know exactly what they need to do, and will build grand things. Growth will be very high. Software will be developed mostly at specialized shops, that require several years of education over the top of a degree. Everything we consume will be not only cheaper and more convenient, but more personalized and tailored to niche interests. It will be … the ultimate fusion of the internet into everyday life, and will probably take another 30–40 years to fully run it’s course.