Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
November 1, 2014
arrowPress Releases
November 1, 2014
PR Newswire
View All

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

Don't monetize like League of Legends, consultant says Exclusive
Don't monetize like  League of Legends , consultant says
August 11, 2014 | By Leigh Alexander

League of Legends has daily active users to rival Instagram and its finals broadcasts see more viewers than Game of Thrones -- across several metrics, it's the most-played game in the world.

But if you're looking at monetization systems, be warned: it's not a good idea to emulate LoL.

Ubisoft Blue Byte's Teut Weidemann has a great deal of experience consulting for online games, studying how they work -- from their metrics to their play experience to their monetization. Using public tracking tools like Comscore, Alexa or TrafficEstimate, Weidemann can crunch numbers that aren't generally publicly available. He researches press releases, conference talks and shareholder reports to get to the meat of metrics matters.

When he plays games for study, he does so at first without paying. "I measure how the game talks to me," he reflects, from measuring the sense of progress and studying the game behavior to friction points and prompts for payment, and he interviews core players. After that, he starts with a small $10 investment -- the most popular entry point -- and maximizes his rewards while buying only the most useful items, and studies how that small investment changes his experience.

League of Legends is fairly open with its numbers, Weidemann says. ComScore and Alexa implement Trojan horses into computers, to monitor users via invisible trackers -- the data isn't always perfect in a vacuum, but can be useful in big-picture studies and comparisons among games.

LoL's players

According to Alexa/ComScore data, most players of LoL are male, and most do not have a college education -- that's because many of them are too young to have earned a degree yet. Officially, over 90 percent of LoL players are male, and 85 percent of them are aged 16-30. Geographic studies and a look at server referrals helps determine the strongest geographic markets for the game (the U.S. and France).

Currently the game has 67 million players, a near doubling just over December 2013. 27 million of those are daily active users; in 2013, that was close to four times as many as the number of users Instagram gets every day.

The game has a dual currency system -- Influence Points, the "soft" currency earned by playing, and the Riot Points, a "hard" currency one pays for. There are rewards for both winning and losing games, and most items require some amount of both types of currency. Players used to be able to earn small amounts of RP through progress, but the company's reduced these giveaways since launch.

Riot releases new champions every couple of months, all available for either IP or RP, with ten of them playable for free each week. Access to a minimum set of champions is important, Weidemann says -- there are hundreds of heroes and players need to learn what each can do in order to be successful. Older champions lose value for ongoing price drops, while new champions are damaged by rebalancing.

LoL's champions

"They release a new champion that is always, always overpowered. So the people who pay for the game buy the champion immediately... and then Riot will go in and slowly devalue the champion until he is 'balanced.' In addition to damaging the goods you just bought, they also lower the price of the previous champion they released," Weidemann says. "I would feel cheated. How come LoL can actually get away with that?"

Riot has freely admitted releasing overpowered characters and adjusting them later, and the player base doesn't seem to mind, even though it makes it hard for players to choose their first champion. The game is very hard to learn organically, and on top of that, LoL's chat, populated with foul-mouthed young boys, was such a deterrent to new users that Riot had to close it.

Riot sells skills for players to customize their favorite champions -- which means until newer players learn about the skins, they might not recognize a champion even if they've already learned his behavior. In fact, even people buying skins (the best-selling item) do not know what the new skin will look like until they purchase it and are forced to research on YouTube. Fundamentally the game is very difficult to learn: "You have no idea what you're doing at first."

Yet LoL's objects for sale are not "pay to win", but mostly cosmetic items, or boosts to the experience of competition rather than the statistics of them. Players love to collect champions, cite enjoyable team play and gratifying PvP. They also find teaming up with friends to be social -- although multiplayer PvP can be inherently anti-social, players share rewards from successes, and participate in a "tribunal system" to deal with problem players together.

Giving too much away?

As of 2013, League of Legends had earned: 70 million registrations since 2009, 32 million MAU, 12 million DAU, 3 million PCU, $1.32 ARPU and $624 million in revenues. The average revenue per paying user on popular games is about $35: Crunch all these numbers and get 1.2 million paying players, a 3.75% conversion rate. That's remarkably poor as rates go.

"Usually, conversion rates for client-based games is between 15 and 25 percent," Weidemann reveals. "World of Tanks has 30 percent. It could afford to have 1/3 of the customer base and have the same amount of money as League of Legends."

LoL gives away too much for free, Weidemann suggests. The company could aggressively increase its revenues across even a smaller player base, and chooses not to.

"Riot doesn't care. Optimizing monetization is not the top priority. They monetize purely through their reach. So it only works because of the large user base, and if you don't have that user base or don't expect to, you should not adopt their monetization. It should not be a role model for your monetization system."

"But you should dig deep and learn why it works, and where the mistakes are," he suggests, "and why they can afford to let the conversion rate be so low. They can afford it; you might not."

Related Jobs

Twisted Pixel Games
Twisted Pixel Games — Austin, Texas, United States

Senior Graphics and Systems Engineer
Twisted Pixel Games
Twisted Pixel Games — Austin, Texas, United States

Mid-level Tools and Systems Engineer
Sega Networks Inc.
Sega Networks Inc. — Madison, Wisconsin, United States

Mobile Game Engineer
Forio — San Francisco, California, United States

Web Application Developer Team Lead


Simon N Tomjanovich
profile image
Maby it works because the monetization is not the top priority. Maby this is one of the major reason of the huge playerbase? Just thinking.

Matt Fischman
profile image
This article contains numerous factual errors from the supposed expert consultant. For example:

"They release a new champion that is always, always overpowered. So the people who pay for the game buy the champion immediately... and then Riot will go in and slowly devalue the champion until he is 'balanced.'

Of the most recent 5 champions in LoL, 3 of them (Yasuo, Lucian, Vel'Koz) were *buffed* in the patch following their release.

I find the rest of Weidemann's statements to be consistently inaccurate and based on broad generalizations that offer little more than a juicy quote.

Robert Carter
profile image
Yeah I stopped reading a little over half way in. Had to reread several sentences thinking "He couldnt have written that" the first time, and then "Has he ever PLAYED this game?" the second time.

I get the feeling he glanced at the game, played to level 10, looked at the forums (where the common complaint is "The new guy is so OP") and then wrote this article.

I havent played since season 3ish (Lucian was the newest character when I stopped), but I played daily since season 1 before that, and this article struck me as being written by someone who has not spent much time in the game.

"Riot has freely admitted releasing overpowered characters and adjusting them later...", I want to know the source of this sentence, as thats a bold statement to make without any source listed. I remember PLENTY of non overpowered characters, even weak characters, being released.

Eric Mickols
profile image
I don't have a citation for you, but I do remember them saying something akin to what the author was saying here about release strength.

Specifically after the release of Caitlyn in 2011. Caitlyn was released with very low damage but high range, and generally viewed as a worthless character upon release. No one purchased, and no one played Caitlyn even after she was buffed in subsequent patches. (Until she became popular in later seasons 2 and 3) Riot had made a comment that they would be leaning towards more powerful/disruptive releases after that, to ensure that the characters made it into circulation before scaling them as harshly for full balance.

Ian Morrison
profile image
I can distinctly remember a few champions (Riven comes to mind) that were actually pretty near useless in the first few weeks of their debut. The assertion is just not true.

It is, however, an extremely persistent player perception in many free to play games. Player cynicism (and a lot of cherry picking of examples) ensures that the player base assumes bad faith on the developer's part. Sometimes it's even true, but in many cases it's not backed up by the facts. Nevertheless, it's a hard allegation to shake because the format doesn't leave the developer above suspicion.

Alex Feng
profile image
Another point I would like to point out is that it's not that Riot Games do not care about their conversion rate. It's always about a balance between revenue per user and retention rate. If Riot Games were to monetise more aggressively, it is very unlikely that they will obtain a similar reach they have right now. How you maximise this depends a lot on gathering data and experimentations.

There are many ways to monetise, you can go wide player base with low ARPU or go focused player base with high ARPU or anything in between. They all make money and there are no right or wrong way.

Michael Joseph
profile image
Indeed. And when you're able to create such a huge installed base, you start to approach World of Warcraft status where many more opportunities open up for extending and capitalizing on the brand.

- models/sculptures/miniatures (collectibles)
- comics
- cartoons
- animated shorts and features
- spinoff titles

and as you say, maybe you don't ever achieve that sort of reach if you aggressively monetize. You also have to look at the lifespan of aggressively monetized games... people burn out from them quickly.

Kenneth Nussbaum
profile image
I think the point is more about how a re-balance can hurt the perceived value of DLC. He goes on to mention that fans/players don't seem to mind or complain about it, I think its to serve as more of a warning for developers who may forget just how intangible the value of digital goods can be.

All in all the data is there, I think the low conversion rate may be due to a design issue. People enjoy it because its fairly balanced, without room for a lot of 'carrot dangling' character growth elements it can be tough to monetize. Johnathan Blow did a recent talk about the relationship between monitization and how it impacts game design similar to how syndicated television and commercial slots can negatively impact and restrict the story arc and plot of a TV show.

Eric Mickols
profile image
I feel like the $6m revenue has to be inaccurate. According to Riot's wikipedia page, they have 9 offices and roughly 1,000 employees. $6m revenue would hardly pay salary in that case, much less their millions in tourney prizes and infrastructure costs.

Maybe the estimate is $6m monthly? Can you explain more what your revenue estimate entails?

Paul Barclay
profile image
$6M monthly ($72M yearly) wouldn't come close to covering the bills. 1,000 employees is above $100M per year in people costs alone. $6M weekly might cover the bills, but my gut says they're making more than $300M per year. $6M daily is too high - unlikely that they're making $2BN/yr

Jeremy Tiner
profile image

Yea, that seemed SUPER off. Does he mean per day?

Christian Nutt
profile image
Right - we've discovered the info wasn't right and updated the story. This story was live-reported based on a talk, and the number must have been mis-transcribed. 6.24 vs. 624. Same number, accidental decimal point addition.

Cristian Vargas
profile image
This sounds like this: "since the moment you buy a champion it will only get weaker". From my experience with the game thats far from being true and even if there are issues with the balance (there are 100+ champs, 4+ game modes and tons of items) they don't just "damage" your goods. He may be right about the fact that riot could make more money, but they always tried to be a costumer based company and if you go to the forums you will see a lot of their employees commited to this end (and their support system is fast and good too from my experience).
Maybe that part of "giving away too much for free" is part of the reason LoL is one of the most played games in the world...i can fight on even ground with a guy that has spent 1000$ in the game, unlike most mmo games (that's one of the top reasons people leaves a game too).
To make that model work you need a huge playerbase to compensate. Dota2 does not seem to have any issue with leaving all of its heroes free and even if their game balance has a different focus you cannot obtain any advantage with money.

Joseph Willmon
profile image
Those are some abysmal metrics. However:

Conversion rate itself is a non-linear metric that does not scale 1:1 with audience size. If World of Tanks had 12x the audience, its conversion rate would be significantly smaller as well.

Monetize your game intelligently in a way that allows it to grow and doesn't sacrifice your vision on the altar of metrics or monetization-mechanics-du-jour. That's what Riot did, and low metrics notwithstanding, they seem to be doing OK.

Andrzej Sekula
profile image
I agree with all the above comments. It's all about the strategy and goals. Friction isn't as obvious as in World of Tanks, where they take rake in form of "maintenance" costs when buying higher tier tanks or when using "gold" ammo when playing CW. As soon as you buy 7+ tier tank, your EV becomes negative (in most cases).

In case of LoL, players start to calculate more deliberately how to balance their spendings after reaching certain level of experience. For example, it's "wiser" to unlock most expensive champions with RP, when they are on sale, for half the price and use soft IP to buy runes, that cannot be unlocked with premium currency.

Just to sum it up, I'd say that LoL monetizes through players' engagement, keeping the monetizing friction at low levels. If the whole business was about ARPPU only, then everybody would make video slots exclusively. Maybe poker and bingo also.

Anders Larsson
profile image
I agree with many of the comments above. I think the stats should be read as follows: $1,35 in revenue per user per month (that is about $70M-$100M per month). This would be more consistent with estimates of last year revenues of about $600M.

His point is however still valid, $1.35 per user per month is not huge, especially as the LoL model actually requires quite a lot of expensive servers and other resources.

WoW still has higher revenues at a fraction of the user base.

Ron Dippold
profile image
He does have one useful (if hopefully obvious) point: You can't expect you're going to be LoL or WoW. If you're counting on that you're probably doomed.

Mike DB
profile image
2013 earnings for League was ~$624 million worldwide. The decimal for "$6.24" is erroneous.

David Keen
profile image
This whole article seems to be saying, "Make your game Pay to Win, unlike LoL, if you want success!" How about we learn the lesson from someone who made a F2P game successful without being (too) exploitative instead?

Tu Nguyen
profile image
I'm really confused, wasn't there an article saying in 2013, Riot's revenue for LoL was $624 million not $6.24 million?

Christian Nutt
profile image
Thanks for the link -- we've corrected the story. This story was live-reported based on a talk, and the number must have been mis-transcribed.

Harold Myles
profile image
Good information but very STUPID conclusion. The one line from the article that highlights its stupidity is:

"World of Tanks has 30 percent. It could afford to have 1/3 of the customer base and have the same amount of money as League of Legends."

That's the point. WoT doesnt have 1/3 the customer base as LoL. Not even close. Why? Because LoL isnt trying to rape their customers or support pay to win.

I am sure if Riot had to choose between their 3.5% conversion rate with 70 million users or a higher conversion rate and 1 million users, they would take what they have now.

Especially when more and more of game revenue is going to come from advertising dollars as time moves on. Those eyeballs are worth alot.

Robert Green
profile image
"That's the point. WoT doesnt have 1/3 the customer base as LoL. Not even close. Why? Because LoL isnt trying to rape their customers or support pay to win." [citation needed]

Thibault Coupart
profile image
Can't agree more with all the comments above.

The statement about champions is fairly wrong, it has certainly happened a few time but it's not a general goal from Riot.

And the analysis of the monetization is really poor as well, the idea to make a free-to-play for core gamer without friction was a really good idea, they stole a lot of audience from Blizzard (which is an exploit, let's be honest),

The lack of precision about how the data is aggregated (monthly / weekly / daily... especially about the multiple quote of ARPU) highlights a poor level of formation in the statistical field. An ARPU is time related, is it daily ARPU, weekly ARPU... and if not it means you are talking of Lifetime value and with this metric you need to take in consideration the retention rate, which is in the case of Lol one of the best in the world.

I am not a fanboy, Riot has some defaults (technical especially), and they tend to throw money a little bit in not-so-good ROI things, but nobody can deny their initial choice of design made their success.

David Paris
profile image
I think the issue that tends to get overlooked, is that this kind of F2P model involves a massive up-front expenditure to compete with, which effectively allows the first couple market entrants to dominate that market thereafter. Sure, LoL may not make a huge average return on user, but the barrier to entry for any potential competitor is _enormous_. All that free content provides an expectation for players on what is available before they even start paying, and as that pool of content grows over time, the barrier becomes higher and higher.

F2P gaming has effectively amplified the advantage of big money backing. As an industry, I think this is pretty terrible for us.

Thibault Coupart
profile image
on the Moba scene, yes that's right. Nearly impossible to compete right now, except if you are Blizzard or Valves.

But I am quite confident that this sort of F2P design can be apply to others gameplay with success.

Sheldon Laframboise
profile image
There is some weird opinions in this article. But the one I find most bizarre is saying that an ARPU of $1.32 is terrible. League of Legends ARPU is in the top 10 of F2P PC games. The number one slot is held by World of Tanks at over 4.51$. Dota 2 is sitting around LoL around $1.32 as well.

They have a proven strategy - and if they changed it, sure their ARPPU/ARPU may go up, but their overall revenue will likely go down. There are certainly some factors I feel you are not considering.

You need to remember this is a community and eSport driven game. Riot/Tencent has built up their community over a long period of time. Due to it's large offering with low barrier to entry their model works because it attracts droves of players. Gating too harshly would destroy their business model and slowly kill off their player base.

Inversely it does take a lot of money up front to make a model like this work. Like all business decisions it isn't without risk and each business needs to pick a strategy that they think will work. I generally feel you are undervaluing the power of their systems design and strategy.

If anyone is interested in a detailed analysis on the League monetization model, feel free to check out my article from last year:

Ian Griffiths
profile image
I don't understand why Weidmann thinks that you can't discount virtual goods as they become less popular, clearly people will pay for exclusivity. We see this everywhere from retail fashion to early access games, whether it's enthusiasm, fandom or social status, some people want stuff before everyone else.

It's a simple notion, but when looking at the monetization for your game, it's important to consider the key motivations behind player spending. Also remember the simple points, ask yourself - 'will my players love this thing enough to actually pay for it?'.

Brutoloco Wilson
profile image
That statistic of Teut Wiedeman is not true.

Check this poll i made on euw forum:

Also everytime i play League of Legends mucho more than the 3,75% of players are using any skin(that menas they paid money for playing)

Totally false accusations and numbers of Teut Weidemann. You can check yourself.

Thibault Coupart
profile image
Interesting initiative;

But the sample is obviously bias, player who read forum are really more engaged than others, especially in Lol because the forum is a bit inactive if you consider the huge user base.

Eric Seufert
profile image
The author of this article neglects to acknowledge that the very reason LoL has achieved a massive user base is because it has not pursued aggressive monetization. He seems to believe that conversion rate and user base size are independent of each other; that is, that altering the product to change one of them could leave the other unchanged (eg. increase conversion rate and keep the user base the same size). This could be true if League of Legends was operating inefficiently, but it's hard to make that argument looking at its longevity and aggregate revenue metrics. More realistically, adding more aggressive monetization mechanics into LoL would almost certainly turn some players off and shrink the user base, which would be objectively bad: even at parity (eg. 5% conversion with 1MM players or 10% conversion with 500k players), the larger player base is always more desirable, as it produces additional virality (among other things).

Thibault Coupart
profile image
and this is where we need real data to state about the topic of monetization vs churn in the case of Lol :)