The changing nature of startup innovation

In these days of rapid-fire technological change, an artificial-intelligence startup can use software to automate a manual process in just three months.

A young AI chip company can generate a prototype in as little as 18 months.

The speed of innovation is faster than ever, it would seem. But is it? Or is the very nature of innovation simply changing?

To many venture capitalists, this question is a critical one as cash rushes to young companies at a record pace. Last year, enthusiastic investors poured $9.3 billion into AI startups alone, a huge jump over 2017.

To some observers, innovation today focuses more on what can be called displacement opportunities than efforts to break new ground with fundamental technologies, with profound implications. If it lowers cost, improves efficiency and steals market share from an existing competitor, tremendous value is created. But if it casts aside fundamental infrastructure breakthroughs in favor of disrupting the business models of big incumbents, that’s another matter.

To some, the best term to describe this less-fundamental innovation is applied, and while it comes with great purpose, it is less likely to open up entire new markets, the way the development of the microprocessor did with the PC market almost four decades ago.

It is a distinction of some importance.

One technology theorist with a thoughtful way of looking at today’s innovation is Geoffrey Moore, the author of “Crossing the Chasm.” Moore, a venture partner at Wildcat Venture Partners, described some of the innovation he sees as “active reimaging,” enabled when technology lowers costs and ends an era of scarcity.

VCJ Venture Innovation
Geoffrey Moore, author of “Crossing the Chasm” and a venture partner at Wildcat Venture Partners.

One foundation for this is cloud computing, which in many ways is the foundational technology today, rather than silicon, and has driven down the cost of software distribution and of delivering compute power and providing infrastructure.

It is similar, in this way, to mobile, which drove down the cost of communicating by phone and increased convenience.

Entrepreneurs have seized on cloud and extended it to new uses. Think of Uber and Lyft, which also capitalized on mobile.

In Moore’s mind, one key extension at present in the enterprise is to what he calls “systems of engagement,” or the move to bring consumer-like Web features to business systems.

A lot of startup innovation has taken place here and more is coming, though in some ways, this trend may have topped out. A lot of the easily reached fruit has already been grasped.

What comes next are “systems of intelligence,” Moore said. With the marginal cost of computing and storage moving toward free, and sensors spreading, massive data collection and powerful computing systems will be able to look for intelligence in a sea of noise.

Google and Facebook already suggest what is possible, as they mine user posts and queries to find valuable clues to behavior.

Applications already target medical research and seek new observations about the human genome and new uses for existing drugs. Sensors tell tech workers when components fail. Commerce will better target consumer desire.

“I think systems of intelligence will be the next mega wave,” Moore said. “I think you’re going to see an enormous amount of innovation there.”

What will follow, in Moore’s thinking, are “systems of autonomy,” which society already has a glimpse of in the driverless car. The real value will be broader, with obvious uses in mass transportation, underwater transportation, mining and elsewhere.

An argument can be made that a lot of this is applied rather than fundamental, though fundamental engineering has taken place. Today’s innovation is more likely to spark displacement purchases and open up replacement markets, rather than double the transistor density of a semiconductor every two years and allow a personal computer.

The purchase of an autonomous car might replace the purchase of a self-steered one.

For some VCs, this is an important distinction.

VCJ Cybersecurity
Robert Ackerman Jr.

“Is innovation alive and well in Silicon Valley?” asked Robert Ackerman Jr., a managing director at Allegis Cyber. “Not the way it once was, in my opinion.”

Ackerman sees fundamental innovation once at the core of improvements in storage and communications being replaced by applied innovation. While today’s innovation is disruptive, its nature has changed.

“What we see is a lot more business process innovation and applied innovation,” Ackerman said.

This may help explain why so much cash has poured into late-stage startups. Financial resources can be necessary when technical barriers to entry aren’t strong enough to protect a business.

Eric Klein, a partner at Lemnos Labs, takes the discussion a step further.

“I think we are at a lull in consumer innovation,” and in particular in hardware, he said.

In hardware, low cost accelerometers, sensors and systems on a chip came along and now companies struggle to find things to do with them that haven’t been done.

“We’ve sort of figured out all the cool and low hanging fruit that came from them,” Klein said.

This is even more difficult when facing competition from big consumer hardware companies, such as Amazon, Apple and Samsung.

“They really do control the distribution points of innovation,” he said.

And often times they are working on a similar product internally. When startups go up against them, they frequently lose.

In a similar way, stagnation may have crept into other once-active areas of technology, too, he noted. In social, for instance, no new competitor has arisen to disrupt Facebook’s business. In enterprise technology, many of the most obvious opportunities unleashed by the current generation of SaaS and cloud companies have been addressed.

Looking forward, Klein sees a different picture. Five massive waves of innovation appear on the horizon and are poised to bring an unusually rich confluence of new technologies to the tech landscape.

They include the human-to-machine interface, including speech and gesture, but also augmented and virtual reality. Alexa is an early sign of what will be possible, and while virtual reality might be overhyped, it appears to be gaining momentum. Also on the list are artificial intelligence, including machine learning, applied robotics, image sensing, and bioengineering and bioinformatics.

With these segments nascent and preparing to take off, “our pace of innovation overall is increasing,” Klein said. “These are all happening at the exact same time.”

But they are not here yet.

Of course, applied innovation is not without value. In fact, there are signs innovation today has continued to generate an expanding amount of value.

Parsa Saljoughian, an investor at IVP, studied tech companies that have gone public in the United States over the past nine years and found that revenue growth at these companies has been increasing. The median trailing-revenue growth rate for companies with offerings between 2014 and 2018 was 20 percent to 30 percent higher than for those from 2010 to 2013, he found.

Several explanations are possible, he suggested. For one, it could mean more mature foundational technologies, such as mobile platforms, open source software, SaaS tools and infrastructure as a service make it is easier for companies to scale. Perhaps as well a higher bar was in place for startups hoping to float stock.

However, it also could point to a higher value placed on innovation. And not simply on the technology used, but the application of the technology.

“It’s the business model enabled by these technologies that is leading to disruption,” Saljoughian said. “It’s not the technology alone. I would characterize that as fundamental innovation.”

Public markets seem to buy into the notion. Last year they rewarded SaaS IPOs with generally strong valuations tied to top-line growth and didn’t trip over negative operating margins.

The private markets are certainly along for the ride. Investors have pushed up round sizes at every investment stage by more than 2x since the start of the decade, according to data from Silicon Valley Bank.

Perhaps one reason for this is the specialized nature of today’s innovation. Take Zinier, a startup with a platform for automating the manual activities of field work. Machines trigger work orders, predict failures and match activities to workers with the right parts and expertise.

Fundamental technology aids the company, which is looking forward to faster 5G mobile internet speeds, along with artificial intelligence and a democratization of IoT interfaces.

But as a result of specialization, the market is “going to see very focused applications,” said co-founder Arka Dhar. “I would say it still is innovation.”

Another element of tomorrow’s innovation is data. Companies have begun to gather great troves of raw data and to enable advanced analytical tools and sometimes machine learning algorithms to find insight. Such a proprietary data set, once cleaned, can provide a formidable barrier to keep competitors out.

When it comes to innovation, reflected Dhar, “I think the pace is actually getting faster.”

To access additional data in PDF form, click on the links below

Artificial intelligence investments soar

Venture deal flow stumbles

Venture funding climbs