By Erin Griffith
On July 12, 2013
For all the excitement around MOOCs — massive open online courses — there is one glaring problem. Yes, it’s wonderful that the Internet can provide access to educational tools to millions of people who wouldn’t normally have access to them. Sites like Coursera and Khan Academy have raised millions in venture capital and grant money, respectively, on the promise that Web-based learning tools will change the world.
Amid that enthusiasm is a stat that’s difficult to ignore: Hardly anyone who starts an online course actually completes it. The retention rates of MOOCs is dismal. There are a variety of studies on this, but the general retention figure I see most cited is a low 10 to 15 percent. Bring on the haters, who condescendingly call any effort to make information and learning tools more freely available “Internet college.”
This problem carries over to the many “learn to code” startups that have emerged in recent years. One of the more prominent and well-funded ones, Codecademy, hasn’t made its retention rates public, but plenty of journalists have piled on to the company’s persistent emails. It is easy to sign up for a free course. But actually learning to write code takes some commitment and willpower to see it through.
The flip side of that is rigorous courses like those offered by General Assembly and Hacker School. They require you to show up and be held accountable. Not to mention, they are expensive. Costs for full-time three month courses at GA can top $10,000. That’s cheaper than, say, going back to college. But you have to be pretty committed to pony up that much money.
So on the high end, you have in-person classrooms. On the other side of the spectrum you have free sites. In the middle is a fairly young startup called Thinkful. Launched last year, the company’s classes are a blend of the MOOC and the traditional school model.
Thinkful classes are not free — they cost $300 a month for a three month course, which is enough to be a commitment but still cheaper than a full-time class with physical classrooms. To keep students motivated and engaged, each student is assigned a mentor who is available for daily discussions and questions. Thinkful currently employs 35 mentors for its first class on front end Web development. Each mentor has between two and 15 students.
The result is a retention rate that is ten times higher than MOOCs, according to CEO and co-founder Darrell Silver. The company measures retention by a “graduation rate,” which requires students to reach proficiency in every topic studied. Between 70 and 85 percent of Thinkful’s students graduate from its front end web development course, compared with a MOOC completion rate of 7 percent. (As I mentioned above, I’ve seen reports of MOOC completion rates hitting 10 to 15 percent, but it is a hotly contested figure and studies vary.)
Silver emphasizes that completing a class is just one step in the process of learning to code, which he views as a two year process. Many of Thinkful’s students have at least started a free MOOC course on Web development and want to invest a little more into their development skills. Only around 10 to 15 percent of them join the program, because they want to land a job afterwards. Silver says many of his students are designers, project managers, or others already in the tech world who want to expand their skill set to writing code.
There are drawbacks to this model, the most obvious being that it doesn’t exactly scale. Thinkful needs to add mentors each time new students join, so it has to grow slowly and deliberately. Currently Thinkful has 159 students, and around 100 have already completed the Web development course. Contrast that with the 200,000 people who signed up for Codecademy in just one week during its “CodeYear” promotion. As Thinkful expands, the company’s biggest challenge will be adding mentors and courses as fast as it wants to add students.
Thinkful has raised $1 million in seed funding from RRE Ventures, Quotidian Ventures and Peter Thiel, who is also an investor in PandoDaily.