In the Marketing BS essays and briefings, I regularly include links to relevant articles from media outlets. In some cases, that content sits behind a paywall. Annoying, I know, but there hasn’t been anything I could do — until now. The Wall Street Journal is testing a new feature that allows subscribers (like me) to share links that go around the paywall. If all goes well, you can now access any of the WSJ articles I reference.
In addition to WSJ, I often cite Bloomberg and the New York Times, two publications that use metered paywalls. As long as you don’t access too many articles from those sites, you should be able to open any of the links included in my pieces. (And if you’re hitting your limit with those outlets, maybe you should consider subscribing?)
Today’s essay is sponsored by
Independent content creators are generating more than $2M per year using this open source, customizable publishing platform. All of the features you need to launch a new blog, newsletter and membership site are built-in — with proper SEO, a clean editor and paid subscriptions with 0% fees.
Building a Roblox Empire
If you are a parent — or you follow tech trends — you’re likely familiar with Roblox. In Roblox’s own words, their core product is “a global platform that brings people together through play.” A recent article in the Wall Street Journal provided a more specific description:
Roblox is a free online platform that features tens of millions of multiplayer games made by its own players with tools the company provides. The games range from obstacle-course challenges and iterations of capture the flag to contests based on popular characters such as Peppa Pig and Sonic the Hedgehog.
The WSJ article also detailed last week’s Roblox news:
[Roblox]’s stock started trading March 10 on the New York Stock Exchange under the symbol RBLX. The shares are trading through a direct listing, bypassing the traditional route of an initial public offering.
Roblox shares opened at $64.50, giving the company a market capitalization of about $42 billion. The reference price for Roblox shares was set at $45, in lieu of a formal IPO price, and is based on recent private-market transactions.
Roblox isn’t a traditional videogame company and is using a nontraditional process to potentially reach investors as the pandemic has driven people to spend more time and money on gameplay.
Today’s essay looks at Roblox in the context of loyalty programs — especially in terms of customer acquisition.
But first, a flashback to my childhood. My family owned a Commodore 64 and I loved playing a videogame called Lode Runner. The premise was simple: your character collected gold pieces and avoided enemies. Lode Runner also featured a “level editor,” where players could build their own levels and challenge friends to play them. (As an eight-year-old, I spent hours creating new levels. I even had a signature style: inserting my friends’ initials into the brick formations).
In videogame history, Lode Runner is notable for popularizing level editors. Games like Minecraft have used building tools as the central element of the playing experience. Roblox has taken the concept of level editors even one step further: in addition to designing levels, players can create their own games.
The Roblox game development tool has some significant limits, particularly with the graphics. Still, players have created an amazing variety of games. A post in PCGamesN provides some illustrative examples of user-generated games: running a theme park, escaping from jail, and adopting a pet.
When I created levels in Lode Runner, I could only share them with my friends. With Roblox, users can upload their games to the cloud, making their content accessible to people all over the world. Plus, the Roblox platform allows independent developers to charge money for their games and levels.
Over the past few years, a robust ecosystem of Roblox developers has emerged. Roblox’s S-1 documents shed some light on the details: almost 1 million developers have earned money from their games. Most of these “engineers” only earned a small amount of money, but like every long tail creator platform, some people have done very well. In 2020, about 250 developers earned more than $100,000.
The company has not shared demographic info about their creators, but the user base of Roblox is very young — more than 67% of users are under the age of 16. The average developer is probably older than the average player; plus, the most successful developers are probably, on average, older than the average developers. In any case, I expect that some of these money-making game-creators are children. A piece in The Observer profiled some high school students who earned up to $4000 — per month — building games on the Roblox platform (too bad monetization options didn’t exist back in my Lode Runner days!).
How does Roblox make money?
Roblox can be downloaded and played for free (which is a common model for many of today’s popular gaming platforms, like Fortnite). But Roblox users can expand their experience by spending money to unlock new games, customize their avatars, and more. Instead of real money, players use “Robux,” the in-game currency. Of course, the (virtual) Robux are connected to (real) US dollars, and that is how Roblox monetizes their platform.
Proprietary currencies can facilitate user transactions, but they also pose some challenges. Let’s take a look at the math behind the games.
Roblox is playable on laptops and desktops, as well as mobile devices.
If a player purchases Robux via their computer, then Roblox keeps 100% of the money (less credit card fees).
If a player purchases Robux via their mobile device, then Apple or Google takes 30% before passing the rest on to Roblox.
A single Robux (Robuck?) costs $0.01 USD. So, even in the worst case (Apple/Google) scenario, Roblox earns 0.7 cents from the sale of every Robux.
Users can buy things that were created directly by Roblox, as well as things built by individual developers. For purchases of Roblox-created digital goods, the company can recognize all of that spend as profit (digital goods are 100% margin).
As for purchases of things created by individual developers, that’s where the economics get really interesting. If a player buys user-generated content, then the creator of that content keeps 70% of the money spent (in Robux, of course). In essence, then, the company is charging a 30% “transaction tax” for players to use the currency.
Roblox also benefits from the structure of currency exchanges. Suppose a developer creates a game that is purchased by multiple users. What happens when the developer wants to convert their Robux into American dollars? Developers can’t just visit Western Union; there is only option for the currency exchange — Roblox itself. And while USD convert to Robux at $0.01, Robux convert back to USD at $0.35.
A simple flow of transactions could look something like this:
A player, on their iPhone, purchases $100 USD worth of Robux. ($100 USD = 10,000 R).
Because the transaction was made via a mobile device, Apple takes $30 USD, leaving Roblox with $70 USD.
The player spends the 10,000 R on a game created by an independent developer. Roblox takes 30%, passing along 7000 R to the developer.
The developer converts the 7000 R back into USD, at the exchange rate of 0.35 cents per R. (7000 x $0.35 = $24.50)
Here is the total distribution of the $100 USD purchase:
$30 to Apple
$24.50 to the developer
$45.50 to Roblox
As you can see, Roblox has designed a system that ensures they receive a substantial cut of all transactions, even when players purchase games from independent developers. Transactions are even sweeter for the significant (but shrinking) percent of purchases that are made via computers, because there is no iOS or Android tax.
Many independent developers are also active players on the Roblox platform. Given the company’s tilted exchange rates, developers have a significant incentive to spend their Robux within the platform, rather than converting them into US dollars. But Roblox still benefits in that case, because they also charge a 30% “tax” for all in-platform purchases.
Suppose a player purchases Robux, which are then passed through three developers before being converted to US dollars. We would end up with a flow of transactions like this:
A player spends 10,000 R on a game created by Developer A. Roblox takes 30%, passing along 7000 R to Developer A.
Developer A spends 7,000 R on a game created by Developer B. Roblox takes 30%, passing along 4900 R to Developer B.
Developer B spends 4900 R on a game created by Developer C. Roblox takes 30%, passing along 3430 R to Developer C.
Developer C converts the 3430 back into USD, at the exchange rate of 0.35 cents per R. (3430 x $0.35 = $12.00).
If the original purchase had been made via a computer, Roblox’s cash flow impact would be $100 – $12 = $88.
But as they say in the infomercials, “That’s not all! There’s more!”
For as long as the prepaid Robux sit inside the Roblox platform, the company enjoys a zero-interest loan — the same way that Starbucks profits from their loyalty program (see “Starbucks Loyalty and Breakage”).
Plus, some Robux will NEVER be converted back into USD, for two reasons:
As Robux circulate through the in-platform economy, the 30% “tax” (applied over and over again) will eventually reduce their value to (nearly) zero.
Some percent of Robux will be abandoned by the user (also known as “breakage”).
The entire Robux system is great for Roblox, but complicated for Roblox accountants. How do you account for the fact that a $100 USD purchase of Robux might: (1) be immediately converted back to $35 USD (minus any Apple/Google tax), or (2) be converted in two months for $24, or (3) never leave the system at all (resulting in 100% profit)?
The accounting treatment ends up being very conservative.
The platform charges (iOS and Android) are all recognized immediately upon transaction, but the revenue is recognized over the next three years. If Roblox is growing (and it is!), this makes their cash flow statements look much, much better than the income statements. Consider this example:
Imagine that Roblox earned an additional $100 in Robux purchases, per month, in 2020 —totalling $1200 over the year. And let’s assume the purchases were made via iOS.
Roblox immediately recognizes a $30/month expense to Apple — $360/year.
Roblox only recognizes $33/year in revenue. Since the increase was spread throughout the year (i.e., $100 per month), the total recognition would be $200.
(We could expect other expenses if users convert Robux back into USD, but let’s ignore that point for now).
So Roblox’s change to their income statement would look something like this:
But change to cash flow would look like this:
And this is roughly what happened for Roblox last year. In 2020, “bookings” (i.e., money paid by users to buy Robux) increased by $782 million to $1.24 billion. During the same time, though, revenue only increased by $239 million (about a third). Meanwhile, the cost of that revenue increased by $209 million, of which $137 million was attributed to Apple/Google fees.
This is the challenge with maintaining an in-house currency. The cash-on-cash impact is generally much, much better than the accounting treatment. That discrepancy is fine when you are building a company, but it will raise questions from investors when you try to explain why your business model is working so well.
Last week, American Airlines raised $7.5 billion by borrowing against the value of their loyalty program. American was not the first airline to take such a step. Last May, United Airlines also raised big money ($5 billion) by borrowing against the value of their MileagePlus program.
In a September 2020 article, Byrne Hobart noted that Delta’s loyalty program was worth $26 billion, United’s was worth $20 billion, and American’s was worth $24 billion. At the same time, those three airlines (as a whole) were valued at $19 billion, $10 billion, and $9 billion, respectively. That would make each airline’s operations worth -$7 billion, -$10 billion, and -$15 billion!
Why are these programs so valuable? If you look at the “name on the tin,” these programs are valuable because they drive airline loyalty. The airlines distribute points, incentivizing travelers to shift their airline spending and/or make incremental trips to earn more points.
But the bigger driver of value is not the points the airlines give away themselves, but the points the airlines sell to third parties. During my time at Expedia, we eventually built our own loyalty program. But before we launched that program, Expedia purchased “CitiPoints” from Citibank. We paid 1.2 cents per point; consumers could redeem the points for a MAXIMUM of a penny. Roughly 40% of the points were not redeemed at all; plus, even when the points were redeemed at the maximum potential, Citi would make 20% or more margin (as affiliate fees on the merchandise offered). When Expedia purchased $100 worth of points, it was costing Citi a maximum of $40 ($100 = $83 worth of points to the consumer, 40% never used, the remaining 60% used with Citi making 20% margin).
Airlines, because of their high fixed but low variable costs, have even lower “effective costs” when points are redeemed.
Which begs the question, “why don’t the airlines just shut down their operations (or spin them off to whoever wants them for $1) and run loyalty programs instead?”
They sometimes do!
Back in 2002, Air Canada spun off its loyalty program (Aeroplan), which was later sold to an investment holding company, Aimia. In 2013, Varig, a Brazilian airline, spun off their loyalty program (Smiles).
Both spin-offs maintained an association with the airlines, but neither did particularly well. In 2017, long-simmering tensions between Air Canada and Aimia boiled over, and AC threatened to cut Aimia from their flight network. One year later, Air Canada partnered with multiple banks to buy Aeroplan back from Aimia. And in December of last year, Gol Linhas Aereas (the new operator of the Varig routes) followed suit and began the process of reabsorbing the Smiles program.
Why are these loyalty programs so valuable when integrated into the airlines, but struggle when they are spun off into separate entities?
Loyalty program spin-offs fail for the same reason that Roblox is not going to spin off their proprietary currency — Robux are only valuable because they are tied to the game. Likewise, airline loyalty programs are only valuable because they are tied to the airline. Most people are content to store their wealth in their home country's currency. People need a very good reason to keep their money in a “proprietary currency” — especially ones that have a history of being devalued (as loyalty points tend to be). One of those “very good reasons” is the belief that the currency provides greater purchasing power than its face value.
Of course, airlines use various tricks to manipulate people’s perception of “value.” Airlines ensure that the value of their loyalty points increases as the absolute balance of points increases. You are probably familiar with this type of situation:
5000 Delta miles are worth approximately $0.
20,000 Delta miles might get you $200 worth of flights (when available…).
But 100,000 Delta miles can buy you a first-class seat worth $4000 or more.
This reward structure encourages points collectors to consolidate their travel spending to fewer airlines. With this model, an airline can shift their commodity product to something more like a “semi-monopoly” (you can only earn AAdvantage points by spending your money with American Airlines).
As an airline expands its network, the value of its loyalty program increases — by appealing to more potential users and by delivering more value per point to their existing customers. If American Airlines creates a new hub in (say) Los Angeles, travelers and potential travelers in Los Angeles will become more likely to value AA points. If American adds more routes from Dallas to other cities, then travelers in Dallas (and in the new cities linked to Dallas) will begin to place more value on AA points.
Consequently, if an airline retains its own loyalty program, then the economics of any given route is the value generated by the route PLUS the incremental value generated for the loyalty point currency. The airline can then, in theory, sell more airline points and charge more for each point it sells.
Concessions and customers
Roblox operates a game-building platform; they also manage a proprietary currency. Improving the quality of the gaming platform does not (directly) make Roblox any more money, but it does increase the value of the currency they sell to consumers.
Airlines are money-losing businesses, but they earn money by selling their proprietary currency (loyalty points). As the airlines improve the quality of their product and network, their earnings will not grow at commensurate rates — they may even lose more money. But, enhancements to airlines’ operations will increase the value of their loyalty programs.
Roblox and airlines are clear examples of this phenomenon, but they are just specific cases of a generalized rule: your business will make different profits on different products, and you sometimes need to offer a low- (or negative-) margin product in order to reach profitability on the attached product.
For instance, Facebook does not make money as a tool to connect friends; but without that service, Facebook would not be able to sell advertising. An even better example: Google runs ads on searches, but the vast majority of their revenue comes from a small subset of search types (shopping, financial, travel, medical, legal). In theory, Google could just specialize on those verticals. If the company tried that strategy, though, users would switch to rival services for their non-core searches. A certain number of those users would not return to Google for the monetizing searches. (Bing considered this tactic at one point. It did not work.)
For another example of “different profits on different products,” think about movie theaters. Chartr visualized the economics of the AMC movie theater chain (pre-COVID):
Imagine AMC theaters only sold tickets to movies, with no concession stands. For the 2019 fiscal year, the company would have earned $3.75 billion in revenue versus $5.07 billion in costs (resulting in an annual loss of $1.32 billion). But in addition to screening movies, AMC is the most profitable restaurant in the world. Their food and beverage services collected $1.7 billion in revenue versus $279 million in costs (for an annual profit of $1.42 billion).
Traditional restaurants operate with a food cost percentage of around 30%. In contrast, AMC’s food cost percentage is only 16% (due to astronomical mark-ups). AMC gets away with high prices because people do not visit their theaters for the primary purpose of buying dinner. Patrons visit the theater to watch a movie; people visit the concessions stand because they sell the only available food and drink items. (AMC has a monopoly on popcorn the way American Airlines has a monopoly on AAdvantage points).
Movies are a customer acquisition tool for selling popcorn. Flights are a customer acquisition tool for credit cards. Roblox is a customer acquisition tool for Robux.
What are you using to acquire customers? And where can you extract your true value?
Keep it simple,
Edward Nevraumont is a Senior Advisor with Warburg Pincus. The former CMO of General Assembly and A Place for Mom, Edward previously worked at Expedia and McKinsey & Company. For more information, including details about his latest book, check out Marketing BS.