The Unlikely Google Killer

I’ve recently seen several articles claiming that Google is a monopoly, is indestructible, and it won’t topple. The Economist recently made similar claims about Amazon and Facebook.

Some of these claims rest on a concept that strategy researchers gave a horrible name to: “time compression dis-economies of scale.” It basically refers to the idea that some resources take a long time to build up, and they are very valuable. In order to gather those resources in a shorter time frame, you’d need a boatload of cash. Or many, many boatloads of cash. More cash than almost any organization could devote to the task.

The immediate example that may come to mind is Bing. Microsoft threw so much money at the problem of getting people to use Bing. They even paid people to search on it. (Insane!) Yet, Bing is a loser. So, we see that these dis-economies of scale are too much even for the mighty Microsoft to overcome with money.

The primary rebuttal to this is that there are lots of little potential Google killers waiting in the wings to take down the giant. Because there are so many, eventually one will succeed.

But Google and Facebook and other tech giants have this habit of buying up promising new startups. It is as if Goliath, rather than fighting David, paid him to switch sides.

If Google continues to buy up any competitors that may eventually lead to its demise, how will anyone kill it?

Blindspot

The idea that Google will devour insurgents by acquiring them is compelling, and an interesting topic in its own right, but it misses a very big point. The thing that will kill Google is something that won’t seem promising to Google. At least not at first.

Will it come from a small firm? Maybe, but evidence shows that major technological disruptions of this nature come from large firms just as often as small firms. (See my previous essay for details on this.) Will it come from Silicon Valley? Perhaps. Will it come from the machine learning crowd? I doubt it. It may come from an unanticipated application of machine learning, perhaps, but Google has its eye squarely focused on machine learning, so it’s likely to intercept threats from that direction. It could possibly come from Amazon, in some form.

The key is that it’s something you and I probably aren’t thinking about right now. Even if it is something you and I are thinking about, we probably haven’t, or won’t, make the connection that it could kill Google until it becomes inevitable.

I think it’s more likely to come from outside Google’s domain of expertise than inside. Since Google is great with automation, Big Data, and machine learning, maybe it will come from a low-tech industry.

Here is a completely nutty narrative, meant for illustration only. Say someone in India realizes that there are large untapped pools of people in her country, and she starts to hire some of them to respond to queries about difficult search problems. Let’s call her startup Insearchant (yes, pun completely intended.) For example, ‘web hosting’ is a really competitive and expensive keyword to advertise on with Google. Suppose that, instead of searching for web hosting providers on Google, a small group starts using Insearchant to find good web hosting. At this point, Google wouldn’t buy Insearchant because it’s totally low-tech. That’s not the future! It’s a step backward. Besides, Google may not even know about this small firm in India. It’s insignificant. But, eventually, Insearchant becomes the default way to search for information whenever the stakes are high. Maybe Insearchant does a better job synthesizing information from all kinds of sources. Over time, more searchers ask Insearchant to find the answer. Google may start to become less profitable, and Insearchant starts to collect more and more data. The trend continues, until, suddenly, Insearchant builds an internal search engine. This engine provides Google-like results, but modifies them according to internal data, data that only Insearchant has. The output of Insearchant’s engine is much better for answering high-value search queries. People start switching away from Google in large numbers. Now, Google makes a mad dash to buy Insearchant, but it’s too late.

Ok, again, I must emphasize that this is only for illustration. I am absolutely not making a prediction that someone will build anything like Insearchant. And even if someone did build something like it, the chances of the story playing out as I describe are essentially nil. The idea was to show that there are millions of ways for Google to die, and many of them could easily look very unthreatening at first.

Conclusion

It may take a while for Google to die (or decay to a shell of itself), but I bet it’ll happen sometime this century. Something will dramatically change the way we collect information, or the way we use it, or the way we value it. We may have alternative interfaces for search, or we may no longer need search.

If you came looking for a specific prediction about who exactly will kill Google, I’m sorry to disappoint.