April 01, 2013

Elena Masolova, EDUSON co-founder, interviewed Daphne Koller, CEO of COURSERA.

- You recently shared data of Coursera user geography and surprisingly Russia was one of the leading countries. Can you share some numbers about Russian users? Which countries are in your focus?

- Our students come from 196 countries and we don’t focus on any specifically. Russia was always in top-5. US is a clear lead by number of students, India is 2nd, while Brazil, Russia, UK, Canada follow. We have about 2.35% of our students from Russia, and we are ranked around 2000 in Russia for web traffic per day (unique users). We would be very interested in talking with top Russian institutions about content for our platform.

- How many employees do you have and what are their roles: website development, video production, signing up new professors?

- We have 40 employees, out of them 23 are engineers who support and develop our platform, 10 people are in course operations team (helping instructors with course production and management). We don’t produce content in house. We have a small business development team who help with revenue models and partnerships.

- You and Andrew are 2 founders, who is the CEO (visionary person) and COO (person who gets things done).  

- We’re co-CEOs and share responsibilities.

- What is your most popular course, how many people have signed up to it? Are there similarities in which courses people like most - is it entrepreneurship, coding, gaming?

- 180,000 people signed up for ‘Think Again: How to Reason and Argue’ by Duke University. 125,000 signed up for ‘Introduction to Finance’ by Michigan University. We haven’t found any pattern to predict course popularity. Initially we thought that more pragmatic courses about career and jobs will be popular, but it’s not so.

- If 100,000 students choose the course then you can’t grade individual assignments and have to rely on multiple choice questions mostly. Essays are not an option (if only not graded by other students). What do you do with this problem?

- First, computer grading is much broader. It includes math expressions, programming assignments, models of physical systems, Excel spreadsheets. Of course we want to have open ended questions, and essays are very important too. For this we’ve introduced peer grading where a student has to grade 5 other works and the result is aggregated. We have detailed how-to instructions, and studies show this is a fairly accurate grading method. This is also part of the learning experience. You see 5 other directions of thought, and get acquainted with different ways of solving the same problem.

- One of Coursera tutors Mark Morley claims[1] that around 3% of Coursera students get to the exam stage and 1,5% pass it. If this number is typical, then clearly you would love to have higher retention numbers. What needs to be done for this?

- Around 5-8% of all enrolled students successfully complete the course. But 40-45% of students who complete their 1st homework pass the final exam. This is fairly high retention. The number is even higher (75%) for those who pay $30-70 for Signature Track (declare in the beginning of the course that they want to get a Verified Certificate). So the overall number is low because people could be exploring and not intending to enroll. We find it very important to create a risk-free environment where everyone can start learning with no fear of failure, as opposed to paying a few hundred dollars upfront for an unknown course.

- Coursera has a non-profit positioning, but you have raised $22MM from reputable venture funds who usually have expectations of 10X return on investment. Can you share some of the plans for monetization? Will all courses stay free?

- Coursera will always offer courses to students for free. We want make quality higher education available to a massive global audience, and in order to maximize access, we believe it is necessary to remove any cost barriers. We are exploring a few monetization strategies: Signature Track, Career Services (successful students can connect with prospective employees, who pay for this opportunity). Some expanding institutions are ready to pay for using our content. In any case we share revenues with our university partners.

- How do you plan to spend the funding? Is it website development, marketing?

- We don’t spend any money on marketing, our growth was entirely organic. It’s word-of-mouth, people share content, recommend to their friends. We’ll spend money on platform development, operations. Amazon Web Services is sending us a pretty large bill for video streaming every month.


- Many new projects in online learning raise venture funding, the competition is going to be tough. What will determine the segment leaders - content quality, website convenience, size of marketing budget, founders’ experience in online business?

- Ultimately it will come down to quality of content. Platform and technology also matter, but great courses from the best instructors are key to success.

- Do you see universities as potential competitors? Will they come online themselves without partners, may be with a goal of promoting the brand (not earning money)? So far they were afraid to cannibalize their offline business.

- Definitely we see universities as partners, not competitors. They have all those amazing lecturers, and we are a platform, an enabler who helps distribute the courses. It is not our expertise to create content, so we complement each other.


- Is it likely that 10 years from now top-level companies (McKinsey or Goldman) will be hiring people without any traditional university degrees, but with certificates of, say, 50 completed online courses?

- Not really. In 10 years a lot can change, but there is tremendous value to a degree. It gives an overview, shows a bigger picture. Online programs’ role will grow, but they won’t replace traditional education.

- Sal Khan is teaching millions of people at KhanAcademy and is, arguably, the worlds' most recognized teacher. Do you think that there'll be much more celebrity teachers? Why do we need thousands of offline mathematics teachers if one person can teach everyone?

- Completely removing personal interaction from learning is bad. There’s so much value to it. Students get inspired by teachers, make career choices because of their mentors. The most effective learning process is a blend of standard online content and a local mentor who knows students individually.  

- Will online learning help developing countries bridge the gap? 

- The only way for developing world to catch up in education and give students XXI century skills is with online services. Traditional means just don’t scale fast and capacity can’t grow as rapidly as required. For example, in India they were talking about building 51,500 new campuses. It’s unrealistic. Coursera provides an alternative solution for dissemination of high-quality content to the people who need it the most. We believe education is a basic human right as opposed to privilege.


[1] http://www.slideshare.net/markuos/mooc-lt-conference-2013-15903824 slide 48

Announcing a new e-learning platform for individual clients: Eduson Academy
Find out more