The (virtual) company that could fix math education and make billions
In this series, Sramana Mitra shares chapters from her book Vision India 2020, that shares 45 interesting ideas for start-up companies with the potential to become billion-dollar enterprises. These articles are written as business fiction, as if we’re in 2020, reflecting back on building these businesses over the previous decade. We hope to spark ideas for building successful start-ups of your own.
In 2004, my husband Dominique Trempont and I started investigating the issue of K–12 education, especially in math and the sciences. Our shared interest in education and our increasing awareness of the decline in the number of American kids pursing science and technology careers was a core impetus.
Dominique had been Steve Jobs’s right-hand executive in turning around NeXT Computer and selling it to Apple, paving Steve’s return to his alma mater. He had also been CEO of Gemplus, a world leader in smartcards, and Kanisa, a knowledge- management software company focused on automating customer service. His experience with education, however, came from being on the board of the International School of the Peninsula for five years.
As part of this endeavor, we did macro-level market research and interviewed a number of teachers and parents at various high schools throughout the Bay Area, deriving direct, experiential feedback from the field. We asked questions about class sizes, skill- gap analysis, teaching methodologies, and supplemental tutoring. At San Francisco’s Galileo High School, Chris Kaegi, a young, bright math teacher, shared a particularly important perspective: “I have a general sense of my students’ skill gaps, but I have 180 students, so even if I know the weak areas, I can’t do anything about it.”
Two core nuggets came out of these interviews: (1) there was no standardized methodology of teaching; and (2) there was no methodology for personalized skill-gap analysis.
Without such methodology, a growing chain of problems arose as a child moved from one grade to the next. A C in seventh grade algebra degraded to a D in eighth grade, followed by an F in ninth grade. If you don’t know how to do fractions, how do you solve quadratic equations?
Lucid was founded upon these foundational blocks, which held implications far beyond local schools and students.
Another scary issue also rose to the forefront. “With an average annual salary of only about $43,000, teaching does not compare favorably with other professional opportunities available to talented individuals,” a 2002 National Education Association research study reported. Less than a third of the teachers and tutors teaching math and sciences had relevant backgrounds. And American schools expected these teachers to come up with the teaching methodology, then execute on it. As if every teacher of algebra or arithmetic was expected to reinvent the wheel!
In India, however, a very large number of students were graduating with advanced degrees in math and sciences, and we saw an opportunity to apply this talent pool to a scalable technology-based teaching methodology that all teachers, all over the world, could use.
First things first: artificial intelligence. I had founded two companies based upon AI principles, and Dominique ran Kanisa, which was based upon the fundamental concept of a knowledge base (a special kind of database for knowledge management, providing the means for the computerized collection, organization, and retrieval of knowledge). With this deep knowledge of AI, we concluded that we needed to align a knowledge base of content with a methodology of teaching. This methodology would include personalized skill-gap analysis, such that a student studying basic algebra could be tested to identify exactly where his or her knowledge gap was. Be it in dealing with fractions or exponents, this knowledge base and analytics software was capable of getting to the heart of the problem.
Furthermore, one of the key findings of our research was that children have different styles of learning. Based on the writings of Howard Gardner, a developmental psychologist from Harvard, we studied the theory of multiple intelligence – kinesthetic, interpersonal, verbal-linguistic, logical-mathematical, naturalistic, intrapersonal, visual- spatial, and musical. Our assessment systems categorized children based on their style of learning, directing them down the right track, such that once their skill gaps were identified, they were taken through the remedial track best aligned with their learning style. A kid with a visual learning style, for example, could be taught fractions in visual diagrams, while a musical child would gain access through musical analogies.
Of course, venture capitalists were none too fond of the education market, since many – including the legendary John Doerr – had tried to penetrate it with only marginal success. Those who did march bravely into the market often languished in an under- capitalized never-never land, unable to take on big problems due to insufficient funding. And so, conventional wisdom followed that you don’t make money in education.
However, in 2008, the Web 2.0 era was fully manifest, and we met Edward Fields, founder of the education venture HotChalk, a free online community application that aimed to connect teachers, students, and parents from kindergarten through grade 12. Unlike many other efforts, HotChalk seemed to be getting real traction.
The brilliance of HotChalk’s strategy? The company didn’t even try to sell to school administrators. They were tapping the core users: teachers, students, and parents. HotChalk was launched in September 2004, and membership had climbed to more than 375,000 teachers by early 2008. More than 7.1 million monthly unique visitors from 188 countries were also using the site. That touched 72,000 schools, 93% of which were in the US.
Against this backdrop, in 2009, we founded Lucid with $8 million in Series A venture capital led by Emergence Capital. It was a PowerPoint financing, with no other asset yet in place. It took us three years of absolute stealth-mode research and development to come up with a scalable methodology for math (arithmetic, algebra, geometry, trigonometry, and calculus) to cover grades 6–12. We then created and licensed an enormous amount of content aligned with the methodology at our development center in India. We involved teachers who had particularly stellar reputations, studied their “art” in great depth, and encapsulated as much as we could into a “science.”
Then came the go-to-market challenge.
Our target market was teachers, students, and parents. In 2008, there were 3.2 million teachers, 55 million students, and over 140,000 public and private schools in the US. Globally, there were 29 million teachers and 464 million students. The English-speaking international market alone included an additional 1.3 million teachers. Beyond that waited the non-English-speaking world, encompassing China, Latin America, the Arab countries, and Europe. The one thing in common: a lack of essential educational resources.
While we wanted this to be a worldwide service that every math teacher at every school in every country adopted to teach every single one of their students, we had to segment the market and find a business model that allowed us to penetrate and get early traction.
We chose North America, partnering with HotChalk. Together, we created a community for middle school and high school math teachers to interact, exchange ideas, and organically engage one another. We also developed a collaborative community of middle school and high school parents at each of the schools our teachers taught in. Most importantly, every teacher who adopted our methodology in their school managed to get the buy-in of the parents to pay for the service. This was very important from a business model perspective since it allowed us to bypass the school systems altogether. However, it also meant that our target customer base remained constrained to affluent families in North America. We did have teachers and families using our service elsewhere in the world, including India, the UK, the Middle East, and Australia, largely due to word of mouth, but by and large, we consciously chose not to fight the battle yet of tackling the less affluent segments of our eventual target audience.
We accepted this segmentation reality for five years, which allowed us to refine our methodology, build company valuation, raise a great deal more financing, and expand into other subjects, including biology, physics, chemistry, world history, world geography, English, and English as a second language.
All the while, we grew our revenues at an average 113% CAGR.
We could have become profitable by 2017, but we took our time. We were addressing a big problem, and we chose to do it right. It involved getting a high degree of personalization in our methodology and the right content to support that level of personalization. The personalization ranged from learning styles to skill-gap analysis to interest-based examples. To a soccer fan, for example, Lucid explained concepts of physics like velocity and acceleration by using metaphors from the field. We also invested in building highly influential content partnerships with the Discovery Channel, A&E, CNN, and Netflix where their content was seamlessly integrated into the flow of our history, art history, geography, and other courses.
We became so well known as an effective K–12 teaching methodology by 2018 that the Gates Foundation came to us with a proposal to fund a rollout of our methodology into poorer schools the world over, providing the missing piece of our intended puzzle.
And that is where we stand today – operating margin of 29% against revenue of $5.6 billion – a world leader in educational technology.