Coursera S-1

As you grow more senior in your career or become an entrepreneur, you transition from functional leadership to business leadership, where you need to deeply understand and execute on the broad business. Reading S-1s is a great way to develop that understanding. Companies reveal significant details of their inner workings, strategies, and financials through their S-1s filings. 

Here I'm using Coursera's S-1 to understand their business, and also general business and accounting concepts. Coursera has both consumer and enterprise customers, so their S-1 gives exposure to both those business models. 

Coursera's Business



Coursera is a two-sided education marketplace. They enable universities to provide and monetize courses and credentials to learners and organizations digitally. For learners, Coursera serves the edtech job to be done of upskilling/promotions, reskilling or earning credentials to get a job, and satiating curiosity and desire to learn. For organizations, they serve the job of upskilling, retaining, and attracting employees.  For universities, I assume the attraction is a new and high-margin stream of revenue, wider reach - more geographies and professionals, and brand-building to attract students to their programs. I personally love their mission to "provide universal access to world-class learning so that anyone, anywhere has the power to transform their lives through learning." I think what Coursera does is close to "unambiguously good", which is rare as most businesses have some negative tensions or externalities. 


Their course catalog includes multiple formats of courses: Guided projects, Courses, Specializations, Certificates, and Bachelor's and Master's degrees. The content and credentials are provided by reputed universities and professors. The quality and authority of creators is its main differentiation from other online learning websites like Udemy or Linkedin Learning. 

Main business Levers 


Coursera has a nice flywheel and network effects, similar to most marketplaces. More and better course content leads to more students, which leads to more paid registrants, which attracts more content creators. The growing user-base of students and universities leads to strong network and brand effects, that create a moat, enable some differentiated learning experiences (peer groups, q&a, etc.), and reduces the cost of acquisition and delivery.  

The main levers to strengthen this flywheel and business are: 
  • Engaged course creators and high-quality, in-demand course catalog and learning experience. 
  • Growing active and paying user-base through acquisition, retention, conversions and win backs to paid courses. 
  • Growing enterprise and subscription sales. 
  • Margin improvements through reducing content cost and through offering higher value and price courses.  

Financial statements 



Revenue

The first line is revenue (Is that why it's called the top line?). Everyone knows revenue - it's the total income a company makes. Well, with a caveat. It's the income earned for the service actually provided already. Say, if Coursera charges a student $12K upfront for a year-long course, then even though they have the $12K in their bank, they only realize the 1/12th of the revenue, $1000, every month. $12000 is the billing or booking amount for the month or year when the transaction happened

Coursera realized $293M in 2020 and $184M in 2019, that's a 59% growth rate. In the prior years, their growth rate was sub 40%, so I'd guess that a lot of the improvement is because of COVID and the key question is whether the growth rate would sustain post-COVID. 

Where does this revenue come from? The breakdown is provided further below in the S-1, in the Key Business metrics section. Consumer revenue is the largest, degrees is the smallest but growing at nearly 100% with higher margins, enterprise revenue is in the middle but growing slower than consumers. 


$192M is from direct sales to end consumers, which I'd assume is mostly from one-time course purchases. Coursera also started an all-you-can-eat subscription package, Coursera Plus, and it has ~50,000 subscribers at ~$400/person/year (same as the price for organizations) - that's $20M or ~10% of bookings. Not bad for year one and I'd assume retention is much higher than one-time course purchasers.  

They have 76M registered users, but that seems more like a vanity metric.  I'd assume most of 46M from prior years are inactive (like me). 


Well, seems like my guess is partially wrong. ~3.6M or 0.5% of the registered user base converted to paying users in 2020.  Looks like half of the revenue in 2020 came from registrants from previous years, which is promising from a churn perspective.  There's a big pool of "life-long learners" and professionals registered over the last decade and there's always some folks in the past cohorts who want to take a new course. Coursera is well set up for re-engagements and win-backs, with new courses and formats, and more effective marketing. 


$70M in enterprise sales to organizations to offer education as a benefit or training to their workforce. They have 387 enterprise accounts, an average of ~$1.8M per enterprise. This revenue is usually recurring and Coursera charges $400/employee, even if they don't use it (in my experience employee education benefits are largely underused in most companies). They are retaining it at 114% YoY. 

And finally, $30M for their degree programs. They have around 26 Bachelor's and Masters' programs, and ~60K students who have enrolled in the last couple of years. 

Costs of Revenue

This is the cost of producing and then distributing the good or service. This doesn't include most staff salaries and sales or marketing. For Coursera, it's $138M in 2020, which is ~47%.  The bulk of their cost of revenue is paying universities that provide the content and the payment varies by offering - 45% for consumer, 31% for enterprise, and 0% for degree programs (seems like degrees revenues are structured as lead gen fees). So if the share of the revenue from enterprise and degree programs grow, their margins will improve. The rest of the cost of revenue is hosting, payment processing, etc.  

Gross Profit 

This is simply Revenue - Cost of Revenue. Coursera's gross profit is $154M of the $293M of revenue, or a 53% gross profit margin. Companies building and selling software usually have ~70% gross profit because selling incremental units of software don't cost much, except for hosting or payment processing. Facebook's is at a whopping 80% and Microsoft's is ~68%. But Coursera is a marketplace and they share 30-50% of their revenue with creators (universities). Netflix has a similar model and its gross profit margin is ~38%. Licensing costs and competition for content is why Netflix is increasingly producing its own content. 

Operating expenses or OPEX 

This covers general costs of running a business like rent, payroll, R&D, sales and marketing, G&A, etc. These are usually fixed costs and not directly connected to each unit product or service. The general hope is OPEX doesn't scale linearly with revenue, so if the company scales revenue, they'd become eventually profitable.  

Coursera's OPEX is $221M in 2020. One strength to note is that Coursera has low marketing and acquisition costs. 84% of learners are acquired organically, through brand, word of mouth, and search. CAC for degree programs is ~$2000, but LTV is >$7500 (their cheapest degree program), so the economics are great. 

Loss from operations

Simply Gross Profit - OPEX. Coursera lost $66M in 2020 and $48M in 2019. Seems like their R&D expenses (~700 people team) and marketing spend have grown. That's typical and even healthy in growing companies, when accompanied by proportional revenue growth, as it shows opportunity and optimism. 

EBITDA

EBITDA is Earnings before interests, taxes, depreciation, and amortization. Coursera has presented an adjusted EBITDA, which also removes stock-based compensation, at minus -$40M in 2020, a lower figure than the loss from operations because of the exclusions. The adjusted EBIDTA margin is -14%.  

Net Loss 

Coursera lost $66M in 2020. Net loss or profit is simply Revenue minus all expense. The main difference from Gross Profits is that the expenses include not just the cost of revenue, but also OPEX, interest expenses, and tax expenses. 

Valuations

April 2019: $100M Series E at ~$1.6B. 11X multiple on 2018 earnings of $140M

July 2020: $130M at $2.5B. 13X multiple on 2019 earnings of $184M. 2020 was a period of accelerated growth for Coursera and hype for edtech so that probably influenced the multiple. 

IPO valuation: Estimated to be $5B. 17X multiple on 2020 revenue. That's a pretty big increase in multiple from 2019, but could be because of improvements in YoY growth rate (from 30% in 2019 to 54% in 2020), improvements in margins, and potentially because the public market and retail valuations are expected to be higher. 

Comps

Pluralsight is an education company that is public. I believe they create their content in-house and they strictly cater to the enterprise IT education market. Only ~30% of Coursera's revenue is from enterprise and Coursera spans beyond just IT education, so this isn't a 1:1 comparison but similar. They are valued at 3B on ~$400M of revenue (8X multiple vs Coursera's 17X). But Pluralsight has a less favorable net margin of -32% compared to Coursera -23%, presumably because of high B2B sales and content production costs vs Coursera's freemium and marketplace model. Pluralsight is also growing slower 23% vs Coursera's 59% in 2020. 

Udemy has both a consumer and enterprise model, so they more similar to Coursera. Their last private round was at a $3.3B valuation and I'd assume they have lower revenue and lower ARPU based on what I know. 

My take

Business and product: Overall, Coursera has a pretty strong and leading position in online education. They have strong network effects through their large learner and university base, and course catalog. 

These days, I'm much more pumped about the new live + cohort-based model of online education, as they are much more engaging and preserve the social aspects of learning. The self-driven, "watch a bunch of videos" model of online education is a more accessible but poorer alternative to classroom learning, with <10% completion rates. But there is nothing preventing Coursera from rolling out that offering at a higher price point and similar margins.  

For corporate education and training, I prefer the model of providing employees an education budget so they can use it as they see fit and a la carte, rather than a specific offering.  Providers then cater to the end-users, rather than to employers, which leads to better learning experiences. 

Valuation and growth: I'm no valuation expert. 17X multiple, if they do IPO at a $5B valuation, seems rich. The public markets are frothy, so who knows. I'm less confident about the 2020 COVID growth boost sustaining post-pandemic, and a drop in growth rates in a half or a year after IPO may not be received well. I'm unsure about the "upskilling" market growing 10X over the next few years as learning is hard and most people are uninterested or don't have the time. But there is a chance that online degree programs can grow 10X from $30M to $300M as they are a cheaper and more convenient alternative to regular degrees. There's a lot of competition for the enterprise market w/ Udemy, Linkedin, Udacity, Pluralsight, etc.