Coursera: Deep Dive Interview with Former Executive
Below is an edited transcript for an interview with a former executive at Coursera. Key issues discussed in this interview include: Why the reach of Coursera matters
- How Coursera acquires content and audience
- The differing motivations of university partners and industry partners
- How Coursera faces multiple competitors on multiple fronts
- Macro headwinds and tailwinds
ARPU!: Can you describe the Coursera platform in your own words, and also the key problems that the users, both from the supply side and the demand side, are looking towards Coursera to solve?
Value Proposition of Coursera
Former Executive: Coursera is a two-sided platform. And it is a way to connect educators with consumers, and so it solves problems for educators, in that it helps educators reach a broad audience.
And even from the beginning, when at first they put a couple of courses on the platform, it was, "Oh, I wonder if people are interested." And immediately, they saw thousands of people interested in this course from Stanford. So the value prop from the very beginning was reach a large and new audience.
You are not just reaching the folks who'll typically take a Stanford course, but you're reaching people all around the world, older people, younger people. And so the value prop for educators, whether it's actually IBM or Stanford, is reach a large, diverse audience.
Why the reach matters
That reach matters for a couple of different things. For educators, sometimes it is about mission. There's still universities that really care about kind of mission and expanding their reach. And that's part of the reason why they exist, so it's kind of reach and social impact and brand, building brand awareness. The second is build a brand awareness for specific purposes. So it might be that it is regeneration for other products that you have. It might be that - I would say the third, and sometimes it's the first, is direct revenue. So there were several partners for whom Coursera actually did deliver significant revenue that contributed to their school's budget.
The purpose for enterprise partners is slightly different. Typically, for enterprise partners, it's more about reach, knowledge building and mission, and typically less about revenue. But that might have changed with kind of some of the smaller partnerships.
It is an audience that otherwise, for most of the partners, would be very difficult to reach on their own, especially because from the start, it was highly global and highly diverse.
Although I say revenue is usually not the top purpose — it can be for some, especially in the beginning. For some folks, they definitely made significant revenue. And sometimes the school takes large cuts, sometimes the actual instructor gets a large piece of it as well. Funny example, there's a guy who created the first series of Python courses on Coursera, and he credited Coursera with fully being able to buy his million dollar house thanks to Coursera, He also has a tattoo with Coursera's symbol on his arm. So he's a full devotee, but there aren't as many like him that have been able to make significant money from Coursera.
I want to understand a bit more on why people come to Coursera. Obviously it is a learning platform, but can you walk through the main motivations?
For B2C users, the pitch is traditionally to upskill, to change your career. I think at some point, there was this idea of you go on LinkedIn to prepare your LinkedIn learning, maybe to prepare for your next presentation. You go to Coursera to prepare for your next job. I think there's a consumer base that just likes to learn. And a lot of it is just they enjoy learning. And I think that's actually still a significant audience that is important. Because if you think about like, a competitor, like Masterclass, which you wouldn't typically think of as a competitor for Coursera, a lot of the audience is mostly just there because they want to become experts, and they just like the idea of learning. So that's still, I would say, significant for the identity of the Coursera consumer.
Now you also have B2B users, and a lot of it also has to do with upskilling. There are two main reasons why a company buys Coursera's product. The first is, it is seen as an important part of a package and a way to differentiate the company. So just saying that they offer learning and training development for their employees, it's an incentive for employees to retain, to join. So it's just seen as a perk.
For other companies, and they were the ones that I think were most eager, it's to train on specific skills, because the company feels like they are behind, and they can't hire fast enough. And so instead of hiring data scientists, they're taking their data analytics folks, and putting them through a data science specializations. And so that's the motivation for maybe a European company that can't fire people easily, but really wants to upskill and is having trouble hiring new talent. We also had a company in India that used it for their new hire. So before their new developers, even like started, during their training they used Coursera products to onboard more effectively. So there's a couple of different B2B use cases for the Coursera content. And Coursera developed the B2B platform, the enterprise side of the platform and the data integration. So to me there's that specific demand, which I think is important, because there's a lot of revenue in B2B. And not all education content providers are able to effectively tap into it. So just a side note there.
And then also a small segment, I think it has definitely grown because of COVID, is universities using Coursera content from other universities to supplement their own education. And used in a kind of a variety of different ways.
Would you say there's any commonly shared characteristic of the user base of Coursera? Is there any sweet spot for Coursera?
A internationally diverse audience base
It's an interesting question. So I would say that the audience is fairly diverse. It's still heavily international. I think there is still an oversized part of the revenue coming from the US, versus an oversized part of the learners coming from outside of the US. I would say that it's a balance typically between wanting to learn soft and hard skills, although there's typically more of a conversion to payments and willingness to pay for hard skills and get a certificate in those as opposed to some of the softer skills. It's really hard to say what the commonality is in terms of income level or like race or even educational background. I think that a lot more of it is for the B2C users. I think a lot more of it is, frankly, about attitude towards education and the need for education, and the willingness to be self-motivated, which is much harder to segment for. But that's kind of the true differentiator.
I want to pick up on the theme of active learners. As far as you can observe, how many of the users are actually active learners in the sense that they will come back to the platform regularly to pick up new skills? Or is it the case that the propensity to learn kind of decreases with age and over time?
Again, this is kind of a very dated snapshot in time. But just to give you an idea, there was a point where only about 50% of those who logged in to Coursera and created an account immediately took a course. Not everyone in the Coursera database becomes an active user immediately. But what was observed is that someone might register for Coursera, then come back months, or even a year later, and then they have a specific course need. They might start a course, never finish it, but then come back for a degree. And so there is a very strong value in old leads, if you were to put it that way, for a learning product, especially if you have really good brand association. And so when Coursera launched a degree product, they were able to tap into a large user base.
LTV of Coursera users
So that's a long winded way of saying, if somebody is not active, it does not mean that they are less valuable of a consumer. Their LTV has a very long time span. Whereas for most products, I would think that you either get money out of them in the first two years, or never, you know. I would say Coursera is a little bit different, because we did see people come out who hadn't been on the platform in a year or two years, when there was new products that Coursera launched.
That being said, obviously, the more active someone is, the more likely they are to enroll in a new product. And if they finish one course, they're more likely to go on to the next course. So completion and being active is still really important.
I'd also say that Coursera only more recently started putting out shorter content. And that's something that we've seen demand for, and that enables a more frequent interaction with the Coursera platform, when you have less time to spare, and so Coursera moved into more of the micro-learning and more of the LinkedIn kind of length of experience, in part to have additional content to actively engage learners when they're not ready for the longer courses. So I think that you are right to ask, when they publicize X millions of learners, how many of those are actually going to create value for the company, but I will say while it decreases over time, there's still a huge value in that overall user base population.
Value of Free Learners
The other thing that I would add to it is the value of free learners. So, sometimes there's this assumption that if somebody comes in, and they have taken a free course and they're only interested in free courses at the time, that they're always going to be a free learner. And that's absolutely not true. And so I think over time, you know, there was a movement away from making free very visible. And now if you go on Coursera's platform - and I mentioned this a little bit - now, if you go on Coursera's platform, you can very clearly see how to take free courses. And they're putting a little bit more money into not just on the product showcase, but even advertising the fact that there's free content. And that's to grow the user base, because that user base is also incredibly valuable, because they can convert to paid courses in the future.
Talk to me on how Coursera look at product and growth.
So the key is really thinking about, who is the customer? How do we think about positioning the product? How do we really think about the growth strategy? So, for example, there was a time when our content acquisition strategy was around making sure that we have all the right content for a set of jobs. So we would do backwards design. So, what are the top jobs that we want to support kind of the learning for? And then what are all the skills that are related to it? Then we kind of switched our strategy to more of a blockbuster model, you know, thinking of how do you mainly focus on getting the top revenue-producing products onto the platform?
How does Coursera add content to the platform?
So I haven't been at Coursera for a while now. But at the time, and I think that this is still true, once someone is a partner of Coursera — Coursera has both university and industry partners — anyone can create content on the platform. That content typically, though, would launch as a course, especially at that time. So courses are about six weeks long. At the time, most courses were free, but had a paid option if you wanted to get a certificate.
But if a partner wanted to launch a bigger product, which is a sequence of courses that we called a specialization, or if they wanted to launch a master track, or if they wanted to launch a degree, that's something that typically we would have more of an acquisition strategy for.
So we would have kind of an outward process where we would ask for specific content, and that process became actually a lot more targeted, as the group developed. And we would have a review process where partners would come to us with proposals and we would review and say, this is something that Coursera is looking to support.
And the reason that there's that difference is because there's a different financial value to a specialization, of which there are fewer in theory, at least initially. A specialization should be unique. So if Coursera asked for a type of specialization, we would say, we are going to help ensure that it's differentiated from other specializations that launch on the platform. And we're going to give it preferential treatment in terms of marketing. And typically, specializations would bring in more money because the groups of courses, they had a slightly different payment model. And so they would also be promoted more heavily in marketing campaigns, in emails, and on the platform itself. And Coursera obviously went through a lot of different platform iterations.
Focus on monetization
There was a period when Coursera went through a focus on monetization, because a couple of years back Coursera was initially fully free. Coursera experimented, very honestly, with certain product design, where it was much harder to see how to enrol in the audit version of the content, especially in a specialization. So for example, for a specialization, if you're on that page, it’s difficult to figure out how to pay, how to audit a course, because you had to go to like a separate page. Still, I would say over 90% of folks coming in on the website overall were able to figure out how to take a free course.
It is still primarily free. But a lot of that has also changed, and there are different types of customers. So, you know, that's why educators partnered to create content on Coursera. And for enterprise partners, some of the reach can be very specific. So for example, for IBM, it was really important that they train folks on certain technologies. I think there's a point where AWS, IBM, Google were trying to compete for experts in the cloud space. They had different programmes that they put out. And so a major mission for Google and for IBM is not just broad training, but to reach developers with X technology and reach Y number of developers.
How do you determine what type of content is in demand and then go out and source for it?
That's a really good question. As I mentioned, once you are a partner, you can organically develop content on the platform. So when I say organically, you have a partnership manager. And you say, like, "Hey, I have a professor of astrophysics that really wants to develop a course." Coursera's like, "Great, you know, work with our team, if you have questions, you have access to the platform, go crazy, we'll answer questions, we'll help you out.”
Evolution of content development
As partners grew and content grew, there was a strategy to do requests or proposals. So Coursera would say, "We're looking for more content in, let's say, project management. Here's the content. Here's the target audience, here are the skills that we're going after." And we would work with our partnerships team to share with all of our partners the list of topics that we wanted, and they would kind of bid on doing these topics. And if they were selected, we would, depending on if they needed advanced financing, give them small advances. And they were typically under $50,000. And this is for specializations. And it ranged, sometimes it was higher, but it was not large advances. And they would be selected and they would create this content.
Later on, we moved to a more targeted way. So it wasn't so much that we would share the list with all partners, but we have a very short list of very targeted content that we wanted to launch, for example, self-driving cars at the University of Toronto, because they had a lot of expertise there.
So that was kind of the process, how do we identify the content to source. Again, that changed over time, because there was a point where we were thinking of even presenting the content on the platform as, first you tell us what your career goal is. And then we would figure out what are all the skills that you need for that career goal, and we would have all of the content to support you.
So the way that we developed the target list was a combination of external of job demand data. And so we would look at what jobs are growing, are hard to sell, and what skills do those jobs require. So to look at externally, what competitors, what type of content they were creating, what are general turns of the market.
Using data to enhance understanding of platform users
The other aspect, obviously, is on-platform data. And so I think Coursera benefited from having a very strong data science team and an ability to collect data on what our on-platform users were searching for, for example, that they didn't find or didn't convert to, what existing content wasn't converting. We could see the background that people had when they were coming onto the platform, and that capability developed over time. So that we could over time see, kind of a more representative sample of, how many folks are coming in who are project managers or developers or, you know, whatever it is. Over time, we would also add questions on What career are you looking to? What's your goals? Are you looking to upskill, reskill, start a new job? And then more data on specifically that career.
We would also do ad-hoc research to better understand specific demands in certain areas, especially once we moved to a model where we were willing to invest more in smaller pieces of content, in fewer pieces of content. So we would do a little bit more research. For example, something like game design has very few jobs, but a lot of people were interested in game design. So that was kind of a potential path to explore. Does that help you answer the question?
Yes, that's an excellent breakdown.
And still, to highlight. Partners can still create courses that they are interested in. And so the majority of courses on Coursera's platform are still created and managed by the partner. And I think that's incredibly important, because it means that Coursera doesn't to be a bottleneck. So the content strategy team can be effective in identifying trends. But there's also a very massive community that is plugged into their own kind of university or enterprise community that's also able to identify trends and potentially add them and I think that's really important.
So for these partners that you mentioned, do you typically need to pay them an advance?
I don't think that this has changed, but typically, it's a revenue-share model. And so this is true for most universities. So I think at most, Coursera sometimes did upfront payments, and that's specifically for universities who didn't have the money to put up content upfront, so like, it was easier for them to...And the way the upfront payments worked is, say Coursera gave you an upfront payment of $50,000 – or $100,000 in like small cases – and it was again, not for courses but for specializations or a series of courses. Before you got your share of the revenue, Coursera would take out the upfront payment before you got your part of the revenue. And so the idea is that Coursera wouldn't lose money on upfront payment unless your content never made over that amount.
What made the revenue share a little bit more complicated was when Coursera launched B2B. I think at the time it was revenue share, but first, take off the cost for the sales team. Because in theory, this was no longer where an audience is just coming to the platform and you're just selling the content, there's the added cost of the sales team. And so that was taken from the top, and then there was a revenue share. I'm not sure how it's fully transmitted to the partners now. But that's essentially how it worked.
And with the subscription model, that it's still, I believe, a revenue share model, but there are some complications, potentially given there's a subscription for some of the content. So you have to figure out how do you allocate the revenue shares based on enrollment and based on different factors. And with enterprise partners, sometimes it's revenue share, but it was typically a smaller part of the revenue, and there were certain partners, especially at the beginning, like Google, that actually did not take any of the revenue. So 100% of the revenue went to Coursera. And in fact, some of the enterprise partners on top of not taking revenue, also paid extra to provide additional marketing support for some of the content that they launched. I believe that they also invested in some translations that Coursera didn't at the time see as a strategic investment to make. But overall, for simplicity it's a revenue share model for all concepts.
Who are the main competitors of Coursera?
So it depends on which product you're looking at. So I would say that for the more traditional, free and short course product: Udemy, Udacity. You could also consider LinkedIn Learning. There are also other short course products. GetSmarter is a competitor, they have kind of slightly longer certificate products.
Boot camps like CodeAcademy are also competitors, because the products can be substitutes. So instead of doing maybe a six month bootcamp, you actually cobble together a series of learning on Coursera, or you go through a series of specializations or a MasterTrack.
And obviously, edX was a more direct competitor to Coursera, especially - and if you think about competitors, it's both on the side of the consumer as well as on the side of the partner. So competing for partners, I would say edX was one of the biggest competitors on that.
Coursera, over the years, has also gone into the degree space. And so in the degree space you are also competing with other OPMs like 2U and Pearson.
Given that it is a fairly crowded field, what would you say are the qualities of Coursera that help the company to win against the competition?
Yeah. So I think it's a couple of things. One is starting with a large customer base. I would say, initially, the focus was building an organic customer base without even knowing what the product and the value prop is going to be at the end. And so from the time they started thinking about monetization, they already had over a million of learners on the platform, right? And so having just a very large scale already creates a competitive advantage, because partners are attracted to the scale, new partners create new content, new content attracts new audiences.
Being able to be a platform that supports the full spectrum of concepts from free to degree, is a huge advantage, because again, the free attracts audiences and their constant acquisition for their more expensive products such as degrees is considerably less than a lot of the competitors and advertising degrees is expensive.
Now, something that's come into debate of whether they can continue having that lower cost of acquisition was increasing competition. And to your point earlier, how much of the customer base can you tap into, but they've been able to grow very significantly over time, kind of proving out that there's still a very large market to tap into both in selling to existing customers to new products, and as well as kind of reaching new audiences. So I think it's honestly that scale continues to create scale.
Coursera is a tech company first
I also think that overall, it's a pretty clean user experience for both the content creators and the consumers. It is a tech company first, it is not an education company first. And I think that that's incredibly important. It's based in Silicon Valley, it attracts a kind of engineering and data science talent. And if you compare it to something like, you know, Pearson, or kind of more of the traditional kind of publishers and education companies that have moved into the tech space, it's very clearly much more of a tech-first product. That enables it to have better user experience, collect data more effectively, and use that data more effectively. That's the other aspect of it.
I would say that they were created with scale in mind from the very beginning. Coursera has always focused on making things highly scalable. And I think that's been really important as just a way to focus on their products. For example, having partners be able to create their own concept is really important to scale, the way that they're able to support the learning experience. So the product itself, I think, allows for sometimes less flexibility on the side of the content creator, at least that's something that I've heard. It's possible that they could have made a lot more changes to it. But at the same time, it's a much cleaner experience. And anyone can really create, you know, if you're a partner, anyone can fairly quickly figure out how to create content with Coursera instead of kind of doing the same sessions over and over again, because they've created a course on how to create courses on Coursera. And they were able to distribute it. And so I think it's like, also just smart things like that have enabled it to move quickly. Having also been at the company, I would say that – and this might have changed as the company grew, but it operated as a tech company and it was fairly lean in a lot of ways, which allowed for faster innovation and also core ideas to be tried, tested, maybe die and then come up again, such as subscriptions, such as a couple of other things.
On the flip side, what do you think is Coursera's most significant risk?
I think that Coursera has a lot of different types of learning products, a lot of different types of partners and has a very lofty mission. I don't know if this is the biggest risk, but I think one of the risks is they have a lot of different competitors in different spaces. For example, for degrees, they have a specific set of competitors; for enterprise products, they have a specific set of competitors. And so just the challenge of meeting the needs of different entities, as well as an organization that purely focuses on that, I think that's both a risk as well as kind of one of the key advantages. When Jeff first started at Coursera, I think he initially was like, "Coursera does a lot. Maybe we should focus on a couple of core things." And then I think two months in, he was like, "Wait a minute, we actually need to double down and focus on everything, because all of it is a bigger system together. Like if you think about it, the universities are both content creators and consumers, the companies are both consumers and creators. And then there was a time when Coursera actually allowed individuals to also create content. I think they did away with that. There's a lot of different things that are happening. That's a challenge.
How quickly they can scale quality degrees is a question, because they've created disruptive degrees. And it is a question of how many universities are out there that are willing to do disruptively priced degrees, that are willing to attract the size of audience they were able to with the University of Illinois? I also think that a lot of US universities are going to be in trouble in the future. And so – and for those that are not – it's also a question of what do they choose to focus on. There's also the possibility that the universities want to keep those higher value products closer to them. So maybe they will not want to put their degrees on Coursera; they'll want to do like maybe one free course on Coursera and work with the ecosystem in that way.
It's also a question of whether they will continue being the platform of choice where kind of all content is hosted or whether because of the revenue share models and because of the cost of creating the content for the partner, there's going to be a shift where universities and potentially even companies are going try to do it themselves.
There's also something like, when Google launched its content, it generated a lot of revenue for Coursera, and it was only available on Coursera's platform. Since then, the Google content is available on Google's platform and a lot of other places. And so there isn't really that many companies like Google who are able to just have have their own platform as successfully. But it's also a question of, is the biggest value in the platform kind, will there be kind of more of a do it yourself approach, if that makes sense?
Thank you for your time.