Technical Insights in Startups

By Ashmeet Sidana

2015

Some of the best Silicon Valley companies have been started based on a technical insight.

technical insight /’teknək(ə)l ‘in,sīt/
noun: If told what, a talented engineer would not know how to build.

Technical insights can give companies an unfair advantage, which is what makes them so important.

Example

One of Google’s original technical insights was PageRank which turned out to be a much better way to rank search results on the web. Like many such insights, it is simple (though not easy!). Notice the straightforward description in Wikipedia: “PageRank works by counting the number and quality of links to a page to determine a rough estimate of how important the website is”. As simple as it is to describe, implementing PageRank at scale is not easy. Their low-cost, highly scalable implementation was another factor that contributed to Google’s early success.

Sometimes technical insights solve known problems. For example, prior to the introduction of special “VT” instructions, the x86 instruction-set was not considered virtualizable. This problem was so well-established that academic papers such as an influential one by Popek and Goldberg were published. However, through clever techniques such as an efficient binary translator, VMware successfully commercialized virtualization software for the x86.

Unfair Advantage

Technical insights are, by definition, novel. They can often be protected by patents. This may give companies a barrier-to-entry by competition and hence an advantage.

Some insights need not be protected by patents. In-fact some of the best ones never are. They can be maintained as trade-secrets and hence do not have to be shared with competitors.  For example, Amazon does not share many of the insights it uses to run AWS. They are trade-secrets of Amazon.

A key advantage technical insights bring to startups comes from the nature of (software) technology itself. Software has low fixed costs, no marginal costs, and is easy to distribute. This makes it ideal for a startup, which must compete against well-established incumbents. This is often why a novel, hardware technical insight is no longer as valuable to startups. They are harder to distribute and protect.

Some attributes to consider when evaluating an insight:

  • High Value: Does it create disruptive value around which a business can be built?

  • Defensibility: How easy is it for a competitor to substitute?

  • Market velocity: Are customers quick to adopt innovation or are they risk-averse?

  • Market structure: Does the market structure allow innovators to reap benefits? It’s hard to make money when selling to an oligopoly.

  • Capital-efficiency: How much capital is required to establish and then grow a reasonable business?

  • Technical risk: What is the risk the technology itself will not work?

  • Scalability/Performance: Is the new solution better? Is it enough to pay for the inevitable disruption the adoption of a new technology entails?

Neither Necessary, nor Sufficient

Technical-Insight-Matrix.png

Startups don’t need to have technical insights to succeed. Facebook, Instagram, WhatsApp, and Twitter are just some recent examples of companies that have done well without a technical insight at their core. Please note that we are not saying that Facebook does not have good technology. Arguably they have some of the best advertising, data-center, machine learning, and mobile UI technologies in the world today. We are just saying that technical insights were not at the core of what made Facebook successful.

A technical insight, however compelling, is also not sufficient. To be valuable the insight must solve a problem with business implications. We see many insights that are clever and novel but do not solve an interesting commercial problem. These insights, more often than not, are better suited to use cases in academia or advancing the state-of-the-art.

An insight must be substantial to become the core idea of a company.  To be valuable to a startup, an insight must also be capital-efficient to implement. Usually startups can only afford a tiny amount of capital to take their product to market.

Do you have a valuable insight? We are often the first investor in companies with interesting technical insights. President/General Eisenhower once said: “Plans are worthless, but planning is everything.” So make your plan, but be prepared to throw it away. In the meantime—when you’re ready—we will be happy to review your plan.

Special thanks to Adina Luo and Balaji Sivasubramanian.

Previous
Previous

The Future of the Data Center