Microsoft Research leaves Silicon Valley: Why big companies aren't startups

VCs often cite how easy it would be for a “Microsoft” or “Google” to come along and crush the little startup that is pitching them, but that rarely happens.

There was a time when Microsoft was a monopoly and would throw together a team to build a product specifically to compete in a newly realized software market, but those days are over. Microsoft is in a buy not build mode these days, and the closing of the Silicon Valley Campus branch of Microsoft Research confirms this will be true for some time to come.

Microsoft has been cutting jobs. In July 2014, it announced it would be laying off 18,000 employees. And on September 18, it laid off 2,100. 160 of those were in California, and the largest group of those in California were from Microsoft Research.

Microsoft Research is the group responsible for innovation rather than progression. The idea behind Research is that some of the best ideas aren’t ready to be tested in products; they need to be fostered by mad scientists so that product managers can come see what they have been playing with and determine which can be integrated or productized.

Microsoft Research is responsible for tablets, large screen touch devices, home automation, smartphones, and more. Microsoft built, tested, and demoed products that looked almost identical to iPad, Nest Thermostats, Siri, Roku, and more. The right founder likely could build startups based entirely on things Microsoft has abandoned, and that is what is wrong with Microsoft today.

Microsoft is not Apple or Google. When people talk about big business and how they are slow to move, or innovate, they are talking about Microsoft and IBM. Microsoft and IBM have always relied on business-to-business sales for the majority of their revenue, and as a result they have been hesitant to build anything they might have to abandon, or that wouldn’t be compatible with the rest of their products. This has always made Microsoft Research a bit of a graveyard for products.

Whereas Apple decides what features need to exist in five years and assigns a team to build those features, Microsoft tends to look at what products might be possible in the future and doesn’t worry about how much those products would cost the consumer. That difference of products as opposed to features has lead Microsoft to come up with some great concepts that no one could afford, and then because there is no viable market, the company shelves the product and it dies.

VCs love to invest in software for a very simple reason. It doesn’t cost anything to sell the second copy. Microsoft Research has done cool software projects, but most of the things that come out of Microsoft Research have been a blend of hardware and software. With no rules on the cost of the build of materials Microsoft often presents solutions like the 50-inch Microsoft Surface, which had a $30,000 price tag. It came out with 11-inch convertible tablet PCs for $2,200 in 2003, when the typical laptop cost half to a third of that.

Many of the cool software products that came out of Microsoft Research also never found a way to be monetized. From “Mouse without Borders” to “Photosynth,” there are all these pretty great products that others would charge money for and make a few million dollars on. But they’re too small for Microsoft to bother with. Without that “bother,” Microsoft doesn’t seem to be able to create new billion-dollar products.

For example, there is a fun little piece of software that came out of Microsoft Research with the rather cumbersome name “Automated video looping with progressive dynamism.” The tool lets you take a video, and it will find five seconds of that video to make a loop of, even if there was no looping in the video in the first place. This is enough of an app to build a “Vine” competitor, or to inspire millions of animated Gifs on Reddit, but at Microsoft it will be an EXE that sits on the company’s server and goes unnoticed (other than the mention in this post and a side note in a patent). It could have been some funky-named dot com: Dynavloop, maybe. Or Autamaloop?

What it really comes down to is that Microsoft doesn’t have what every startup that succeeds has: business sense. VCs don’t invest in a really cool technology, they invest in a large potential return. When you point engineers and people with doctorates in various fields at a problem, you may get a result, but that doesn’t mean you’ll get a business.

Brandon Wirtz is a former Microsoft employee who now works with startups to grow their reach, develop their products, and determine their place in the competitive landscape as CEO of BlackWaterOps.com.