How Harvard Business School Is Reimagining Online Education

The control room for HBX Live! at public broadcaster WGBH

Byrne:    I was shocked at how low the price was, because it’s probably the least anyone could pay to have Harvard Business School on their resume.

Anand:   Yeah, and at that time it was $1,500 more than Wharton. People were asking: “How can you charge $1,500, when Wharton’s free?” And we do regard Wharton as a peer. But Wharton’s not our competition in this space. We are trying to create an experience for a learner that is highly curated, and holds your hand, from beginning to end.

Byrne:   In June of 2014 when the beta launched, you made it available to undergraduates and alumni of colleges in Massachusetts and then followed up with the B2B test. What did you discover initially and how has it evolved?

Anand:   About 15% to 30% of our students come here without deep knowledge of business. We’ve got to evaluate: “Are they ready for day one of the MBA program?” So we started by creating three pre-MBA courses: financial accounting, business analytics and e-commerce for managers. We created these three courses online, what we call CORe: the Credential of Readiness. The idea was to create a 10-to-11-week program for students to get up to speed when they come to campus.

There was no reason to keep CORe limited to our students. In fact, the first cohort was not our students. It, as you said, was a pilot, with students restricted to Massachusetts colleges and universities. The logic there was if something goes wrong, we can reach out to them. 

We had about 600 students in that first cohort, and we learned many lessons. The first was sometime in April, just after we had announced the launch of HBX. We were sort of twiddling our thumbs, waiting for applications to come in. Then we realized we didn’t get the word out. We had a “silent marketing strategy.” We got one application, and he wasn’t even eligible, because he was from San Diego.

Byrne: So much for notion, “Build it and they will come.”

Anand: Indeed, I learned three big lessons: The first thing is about engagement. On the first day we started getting feedback from learners saying, “This is unlike anything I’ve taken online as a student.” One Harvard biology undergrad logged in at noon that first day and finished the first module of all three courses by 9:00 p.m.

Obviously, that is really quick, and she seems pretty smart, but had we gotten it wrong? I got an email from her, saying “Professor Anand, I just wanted let you know I’ve viewed the platform for the last nine hours at a stretch. I can’t tear myself away.”

It was one of those amazing moments where, you know, this is a year-and-a-half effort, and we got it right. But we still had problems. We released the modules over time and had encouraged students to do a little every day. Of course, when the module deadlines came along, everyone tried to finish the modules at the same time, and the systems crashed. So we had to make those more scalable and robust.

We were looking at two metrics from that first cohort: engagement and completion rates. Completion rates ended up being about 83% to 85%. What’s interesting is we’ve built the platform so we had no live faculty interaction with the students. I think this is really important as a design principle. Online education right now involves a trade-off between reach and engagement.

You can broadcast a TED talk or a large online course and reach millions of learners. Typically, those (reach) models have low engagement. Faculty conversations with small groups of learners are fantastic for engagement, but not so good for scale, because the faculty is a bottleneck. How do you break that trade-off? For us, the design principle was highly engaging outcomes with no live faculty. Otherwise we won’t scale. How do you achieve that? With social learning?

We were relying on the learners to go to the discussion boards and help each other with their questions. We thought about searchability, fun, and incentives, and we got 75% to use the discussion boards. We worried that they wouldn’t answer the questions the right way, so we had content experts ready to jump in if the discussions went wrong. The number of times we had to jump in during the first three weeks was zero.

Byrne: In fact, wasn’t it even difficult to resist jumping in?

Anand: Very difficult. One of our most experienced faculty told us one line when I came here 20 years ago. He said: “Trust the students.” Your instinct is to jump in and correct them. Let students discuss with each other, and they’ll figure it out.

It’s even harder online, because you’re not seeing the students. You see a comment which may not be right, and you feel like jumping in. But we waited and sure enough, the conversation converges. It’s very humbling. They don’t need us. One of the first people to take the course was the head TA for computer science 50 (CS50), which is the most popular MOOC at Harvard. It has 300,000 students and 100 TAs, so that every time you answer a question, within five minutes, a TA jumps in.

He said, “Your design principles were literally the mirror image of CS50 in every respect. Completely the opposite. He didn’t realize that so many TAs, while very effective in one sense,might also be reducing the incentive for people to answer each other’s questions.

Byrne: Actually, that’s a big surprise, because you often have undergraduates with no full-time work experience. They can’t take what they’ve learned from the workplace into the classroom.   

Anand: That’s right.The first cohort was fairly young, and they did really well. We thought they were an exception because they were from Massachusetts, which has some great universities, colleges and kids. But we opened CORe globally six months later to anyone, anywhere, and got 900 people from 300 universities, who did better than that first cohort.

Byrne: Were the completion rates as high as 83% to 85%?

Anand: Completion rates have stayed steady now for about … 19 waves or more. They remain roughly 83%. It starts shaking our assumptions about the distribution of talent and where it lies. 

When we opened up globally, we said we’d restrict CORe, a pre-MBA program, to people three, five, and seven years out. A third of the applications we received were from people who were over 30 years old. Our first instinct was to reject them. Then we started looking at the applications, and this was a big learn. We got people from all walks of life, physicians, lawyers, educators, airline pilots, real estate developers, software engineers, school teachers. And the age distribution spans 18 to 60 years old. They all said they never learned this stuff before.

Byrne: That’s fascinating and rewarding.

Anand: We got rid of that residential boot camp, which was called analytics and replaced it with CORe, and required 150 would-be MBAs to take it before going into the first-year curriculum at HBS. We decided to open it up to the whole MBA class, because there might be other students who couldn’t get into the residential boot camp because on paper, they didn’t need it.

Our admissions director was supportive, but skeptical: “No student will spend 10 weeks in the summer on an online program before they come to campus,” she said. Anyway, we gave them the option. That first year, 350 entering MBA students took CORe; an extra 200 beyond those required to take it. A third of our MBA class now takes CORe, in some form, before they come here. We run into students who say, “Thank you for creating CORe. I was a banker, worked in Wall Street for four years. I’ve taken finance. I thought I knew this stuff. I actually didn’t.”

Byrne: So you can use CORe to replace the old residential boot camp to even up incoming students.

Anand: This is why we created CORe. You might have taken four years of economics or two years of accounting and you still don’t understand the fundamental concepts deeply. We have a program for leadership development (PLD), for people who have, say, seven-to-10 years of experience. The head of this program wanted one of the CORe courses as module zero for that program because he didn’t think they understood the material.

I thought he was stretching this. These are executives, who have been out in the world for maybe 10 years. But they loved it. Then the faculty chair of the General Management Program (GMP) program wanted it and now we’re talking with 15-to-20 years of experience. Finally, for the advanced management program, which is our senior-most executive program with people who have 30 years of experience, wanted it. They’re now taking the financial accounting course before coming to AMP.

Questions about this article? Email us or leave a comment below.