What’s a tech company, anyway?

Talk of the "tech sector" is out of date. Every company is a tech company.

Uber has encountered a series of challenges that are notionally unfamiliar to the current generation of tech companies: wrongful-death lawsuits, rent-seeking by an entrenched industry, regulatory scrutiny from local bureaucrats, worker protests. The company admitted to having disrupted a competitor’s operations by calling its cars, then canceling. No matter how explicitly it warns about surge pricing, riders accustomed to a certain way of booking a car ride object.

There’s an established industry that charges people for rides in cars, and it’s been reduced to a set of straightforward points of competition: price, car quality, ease of booking, and — treacherously for Uber and uncharacteristically for “tech companies” in general — the burly and distasteful accumulation of political clout before municipal taxi commissions.

It is a testament to the strength of the tech industry mythos that anyone is surprised this is happening. Uber is a car service with great dispatching software, and it’s encountering car-service problems. Yes, it’s based in San Francisco, has been funded by Google, and employs an army of software engineers, but it’s not that much more of a tech company than John Deere, whose tractors are loaded with sophisticated software [paywall].

My own writing and speech is peppered with lazy and imprecise references to the “tech sector” or the “tech industry,” which are either absent or hazily defined in the formal taxonomy of economists and government record keepers. Those terms are convenient shorthand for a kind of research-driven company that we know when we see. In practice they get applied to any company that deals with software, provides services over the Internet, or develops electronics. That’s an absurdly broad characterization, and companies and politicians take advantage of it to suggest intelligence, novelty, and high growth ahead.

The nebulous classification means we tend to call companies “tech” when they approach any kind of problem with a software mindset — the idea that some computer intelligence could add value to something we already do.

There are several widely-accepted ways to classify companies by industry, each of which starts with a sector — a broad part of the economy like agriculture, mining, finance, transportation, and so forth — and then branches out into industry groups and industries. Google, the archetype of a tech company, falls under Standard Industrial Classification code 7370, or “Services — computer programming, data processing, etc.,” part of the service sector of the U.S. economy and the business services industry group.

Also in that classification: American Tonerserve, a wholesaler of printer cartridges; Gray Fox Petroleum, an oil prospector and provider of architectural renderings to the oil industry; and Vape Holdings, which markets e-cigarettes and marijuana accessories.

In the somewhat more modern North American Industrial Classification System, Google is a member of industry 519130: “Internet publishing and broadcasting and web search portals,” part of the drilldown from the information sector (which also includes newspapers, recording studios, and libraries), and then through the “other information services” industry group.

Google’s companions in this classification include all those websites that, yes, are on the web, but that aren’t really anybody’s idea of high technology: “Almanac publishers, exclusively on Internet”; “Greeting card publishers, exclusively on Internet”; and “Special interest portals (e.g., parents sharing information about child rearing, etc.), Internet.” If you can append a comma and “Internet” to your business, you’re about as close to a tech company as you can get.

It’s only in the Thomson Reuters Business Classification [pdf] that we have a technology sector, and it’s also too broad to capture what we really mean when we talk about “tech companies.” Here, Google is part of the technology sector, but so are long distance telephone companies, wireless carriers, software wholesalers, and payroll handlers.

This isn’t intended as a takedown of classification efforts — they’re useful, but only if you understand what goes into them and interpret them with restraint. Tech companies are so hot right now that every company wants to be one.

Is Groupon a tech company? It employs a lot of software developers, provides services over the Internet, and has an app. So does Deere & Co., but we call it a farm machinery and equipment manufacturer. Boeing has more than 40 openings for software engineers right now, many in Google’s hometown of Mountain View, Calif., but we call it an aircraft manufacturer. Toyota’s reputation has suffered lately over software glitches, but, of course, we call it an automaker. Let’s call Groupon an advertising agency.

The concept of a “tech sector” is out of date, and we should stop talking about it. That doesn’t mean the idea it stands for is over — in fact, it’s stronger than ever. The tech mindset has permeated every other part of the economy, and construction firms, heavy manufacturers, transportation services, and insurance companies have basically become tech companies.

Our upcoming Solid conference is a recognition of that. Back in November, as we were starting to put the conference program together, I wrote that every company is a software company. As software reaches beyond the confines of monitor and keyboard, it becomes a bigger part of every business: a way to add intelligence to things that haven’t been intelligent before; a way to instrument and optimize physical machines the way that web advertising and financial portfolios have been instrumented and optimized over the last 10 years.

I meant this in two senses: first, that every company must now understand software in order to survive, and second, that as software and hardware collide, we’ll see a lot of big advances come out of companies you wouldn’t think of as tech companies. Manufacturers, airlines, and electrical utilities all have incredible expertise in what they do, developed from operating in very difficult, highly evolved industries. They understand the first principles in production engineering, queueing, and power distribution, and given the right ways to interact with the rest of the software community, they’ll be able to develop powerful platforms that can attract other companies, large and small. We can see this emerging through companies like FedEx, GE, and Ford, and the question of how partnerships will develop between software companies and builders of things will be central at Solid. The tech sector is wide open.

tags: , , ,