Growing the innovation economy with precision and scale

The need for a scalable way to find and fund impactful startups

Global entrepreneurship continues to grow at a very rapid pace. In 2017 $140 billion of venture funding was invested globally. $2.3 trillion of value was created in the startup economy between 2015 and 2017. [1]

One substantial challenge to continue this growth is access to smart capital. Crowd funding and cryptocurrency are rapidly growing attempts to create a more level playing field for capital access. However, these attempts fail on two critical startup success factors:

  1. Ecosystems and technologies that develop entrepreneurial skills and help founders iterate ideas into a viable going concern, secure resources and grow businesses.
  2. Technologies that enable startups to identify market opportunities and investors to find and invest in successful companies, both with predictive precision and skill.

Silicon Valley has led the world in value creation through entrepreneurship over the past several decades. I write this piece with the perspective on entrepreneurship that resulted from 36 years as part of this community.

It started here in the San Francisco Bay Area

I came to the San Francisco Bay Area in 1982 to join with Professor Ed Feigenbaum and three other Stanford University Professors to lead technology development for an AI company recently founded. IntelliCorp, where I later became CEO, went public at the end of 1983 as the first publicly traded AI company. During the 80s we created and introduced expert systems that are now part of the economic infrastructure: value-based pricing of airline tickets, simulation and scheduling of complex manufacturing lines, configuring complex systems (trucks, computers, construction equipment), and advanced financial modelling, to name a few. Following IntelliCorp, I was CEO of Connect, which went public in 1996 and led the first wave of large scale ecommerce, and CEO of Informative, which was an early leader in corporate use of social marketing.

During these years of raising money and building companies I often had ideas of how to make it easier for the next entrepreneurs.

Taking it to the world: creating a level playing field that predicts stars

Over four years ago I began working on a project with co-founders Fred Campbell, Markus Guehrs and Kim Polese to create a rational and predictive platform for helping early-stage startups improve odds of success. Our goal was to create a process and platform that would accurately pick the stars while creating a level playing field for both market information and capital. We have been scoring and investing now for 28 months with some interesting outcomes: 1. the predictions track extremely well to funding momentum (80%% have raised follow-on financing at increased valuations ), and 2. Over 42% of the companies are founded and run by women (over 56% by women and minorities). This later result came about because we had a scientific process. The company is CrowdSmart. Our Chairman, Kim Polese, will be speaking at DLD19 conference in Munich, Monday January 21, 2019.

A Platform for Prediction and Learning

Making the right decision is hard for any investment, particularly so when investing in early stage companies where data is scarce and the unknowns far outweigh the knowns. The typical process for entrepreneurs is to find access to some funding from family, friends, or other relationship-based approaches such as angel investment groups. There is little to no market data collection or organization of evidence for or against an investment. As a result only a small fraction of these companies are successful. (Specifically only 10% of all early seed funded companies get Series A funding, and on average only 27% of series A companies have a profitable exit. Less than 2% of the of the billions invested in seed startups generate the big returns that make news headlines. The truth is, the vast majority of seed capital invested is completely lost) ) This inefficient use of capital needs radical improvement. However, keep in mind that it is this noisy process that has led to the tremendous results we have had so far. But we can do better. In addition, capital is often denied to women and minorities and also unevenly distributed geographically due to the relationship nature of early-stage funding.

Updating beliefs based on evidence (Bayesian Learning) is fundamental to making the right investment decisions particularly with early-stage companies where judgment and imagination are needed. Furthermore, multiple perspectives from diverse backgrounds and experiences reduce errors in judgment. For this reason we (CrowdSmart) built a platform that is highly versatile for “Collective Bayesian Learning”. allowing groups of people to consider the evidence presented by a startup, interact with the startup team and ultimately produce a body of quantitative and qualitative (natural language) data that provides a foundation for predicting whether or not a company will gain funding momentum and go on to a successful return for investors. Companies that do not score well can learn from the feedback and try again,. The CrowdSmart Collective Bayesian Learning Platform is designed to create the most accurate and scalable methodology for predicting investment outcomes. Because the system continuously learns with each investment, the predictive accuracy and support knowledge base grows in accuracy and intelligence over time. As mentioned earlier, on preliminary initial data, our current conversion rates to Series A are 80% vs. the industry average of less than 10%[2].

There is a substantial amount of complexity and detail in the underlying platform, but from the investor and startup point of view it is simply a way to create a scientifically based predictive process that delivers reliable results. Both the startup and the investor get to see the data as it evolves, providing a learning environment for both parties.

A Simulated Learning Lab for Entrepreneurs

When we first formed the company, the initial vision was to connect with entrepreneurial initiatives in top universities to use the technology to create a learning lab so that entrepreneurs could address key issues like product market fit or market need. University ecosystems drive a massive amount of the value creation in new companies worldwide.

We quickly learned that one of the biggest issues for startups was transparency in the investment process. Entrepreneurs needed to know what it really takes to create a fundable company with a high to reasonable probability of success. Thus we pivoted to an investment model. We decided to fund companies based on achieving high enough scores along key dimensions such as market need, product market fit and team quality. Over time we developed a trained model able to predict at a high level of accuracy those companies that have the highest probability of a profitable exit.

We are now exploring ways to enhance the work of accelerators, universities, and other organizations to foster successful entrepreneurs by providing a learning laboratory that is built on real investment knowledge and best practices. Because it is based in a knowledge-based, machine learning technology base, the platform is scalable to address the worldwide needs for training entrepreneurs to fuel growth of the innovation economy.

The value of a metric

There is an interesting parallel here if you look at the mortgage loan business. Early on, home buyers qualified for a home loan based on their relationship to their local bank. It was an artisan business built on personal relationships and community. For that reason, it did not scale well.

In the 50s an engineer (Bill Fair — seriously that was his name) and a mathematician (Earl Isaac) created a score that ultimately opened up the mortgage lending business to a nationwide market, creating a somewhat level playing field for getting a mortgage. While it took time to take hold, an entire industry formed around the FICO score. Credit Karma has a $4 Billion market cap based on helping borrowers improve their credit score.

Our goal is to establish the most predictive score for early stage startup investing, level the playing field, and create access to capital while also selecting the most likely companies to achieve success, have a positive impact and return a profit to their investors.

Much of what also drives me personally, and my co-founders as well, is the passion that this will open the doors of opportunity and capital to the most creative and brightest innovators world wide. Our mission has been and continues to be “to significantly increase the success rate of startups.”

Taking it global

From the beginning our mission has been global. We believe there is far more talent than there is opportunity. People in the developing world could easily be the most innovative entrepreneurs in the next decade or century. People closest to the problems experienced in the major population centers of the world will have the best ideas of how to meet those needs. If we believe that idea generation may be the highest outside of the US. but that actual development of organizations will not happen without capital and mentoring, the opportunity for creating a metric and a level playing field is likely to have far greater impact globally than in what we have seen to date in Silicon Valley or in the US. This is why we created CrowdSmart.

[1] Global Startup Ecosystem Report 2018 (Startup Genome Project)

[2] VC backed startups have a 35% conversion rate to Series A and the very top accelerators have a 48% conversion rate to Series A. However, there are limits to the scalability in VC partnerships and accelerators.

--

--

I love pioneering transformative technologies based on solid science. Co-founder and Chief Scientist at CrowdSmart.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Tom Kehler

I love pioneering transformative technologies based on solid science. Co-founder and Chief Scientist at CrowdSmart.