Category Archives: Games Scratching, Mixing, and Spinning Out a Business Model has lifted off with ferocity, addicting those with early access while motivating others to find a way through the velvet ropes into the virtual club rooms.

What characteristics of proved to be the booster fuel, enabling this type of launch buzz and engagement?

  • Many of us secretly or not so secretly, have a desire to be DJs, so this concept for sharing music is compelling: a context where the DJ and other users are anticipating what song will be pulled out of the milk crate and whether the selection will drive forward the ongoing mix.
  • There is a gaming element (winning points gains you better avatars).
  • There are social elements (get your friends and work colleagues in, along with meeting new people based on musical taste).
  • Related to both the social and gaming aspects is the quest for social approval for the user’s mix skillz and general musical taste.
  • The combination results in intense engagement providing an opportunity for music discovery, giving at least a glimmer of hope for a business model to be built on this. (If the labels cooperate, historically not a given, so this is a big if.)

Any lessons here for other concepts?  Something sports-related could share these characteristics with the right implementation. Perhaps something around fantasy sports and real-time play calling during live sporting events.

The question with all rocket launches is whether the rocket reaches orbit or does it come crashing back to earth?  If we are still talking about a year from now, and we very well may be based on early returns, we will have some more lessons to draw about keeping something from being a passing fad and creating something more durable.

Zynga S1: Revenue Model II: Getting Users to the Virtual Cash Register

My first post about Zynga’s revenue model concerned the number of overall users.  Given the performance over the last year, let’s assume for a second that future growth of Zynga will not be driven by the overall user variable.  With that assumption, more pressure is put on the variables of the revenue model that concern actual spenders; namely the percentage of overall users who buy virtual goods and the average amount that they spend.

There is little direct evidence on the value of both these variables in the S1 as far as I can see.  However, the financial performance over the last year shows that one or both of these variables have been on the upswing.  As noted in the table below, revenues, bookings, and adjusted EBITDA have gone up on a quarter by quarter basis, despite the fact that as noted in my prior post the overall user number significantly dropped over some of this period.  So, for example, average DAU went down 20% from June 30, 2010 to December 31, 2010, but revenues went up approximately 50% and bookings went up about 25%.

This is an incredible and significant performance for Mark Pincus and management; even more so, when one considers that at some point in this period that Zynga took a revenue hit from the Facebook Credits transition.

That said, from the S1, I have no sense of how much more these two variables can grow.  How much scope does Zynga have to incent users over time to pay for virtual goods, and how much scope is there to get existing buyers of virtual goods to spend more?  These are significant open questions.

9/30/09 12/31/09 3/31/10 6/30/10 9/30/10 12/31/10 3/31/11
















Adjusted EBITDA








Zynga S1: Revenue Model I: Overall Users? Fade and Refresh Pattern

The Zynga revenue model, dependent as it on the sale of virtual goods, is driven by the following variables:

  • Number of overall users,
  • Percentage of overall users who buy virtual goods, and
  • Amount spent per spending users.

This post looks at the first variable: number of overall users.  There is data in the Zynga S-1 for the last year and a half.  The general pattern is that the various measures of users (all of which overstate actual unique users, see S-1 for definitions), peaked in 1Q of last year, dipped for the balance of 2010, and rebounded to approximately the 1Q 2010 peak at the end of 1Q 2011.  One of the factors listed for the rebound this quarter is the launch of Cityville in December 2010 supplemented by adding content to existing games and the launch of certain mobile initiatives.  The chart below has the average daily average users, monthly average users, and monthly unique users.

9/30/09 12/31/09 3/31/10 6/30/10 9/30/10 12/31/10 3/31/11
Average DAUs








Average MAUs








Average MUUs








The question is whether there is much growth left in the number of overall users.  Is there an upper ceiling?  The short answer is that we don’t know, but one hypothesis is that the most recent numbers are not far off from the peak of overall usage, given the numbers have peaked at about the same levels twice.

Another related possibility is that the pattern in users is one that fades – represented by the drop from the period from 3/31/10 to 12/31/10 – and which needs to be refreshed by the release of hit games.  It appears on an aggregate level that usage drops until something new is introduced to refresh interest.  This is also clear from an individual game level, as shown in the table below which combines disclosure of individual game MAUs as of June 2011 in the S-1 and all-time peak MAU as available on the Appdata website:

MAU June 29. 2011 MAU Peak (AppData)
Cityville 70 million 101 million
FarmVille 37 million 84 million
Zynga Poker 35 million 38 million
Empires & Allie 27 million 45 million
Frontierville 11 million 37 million
Mafia Wars 8 million 28 million

One would expect that interest in individual games would fade over time — think Pacman, Donkey Kong, or any other game you have ever played. Consistently, the MAUs for individual Zynga games show a peak and a deterioration over time.  (It would be interesting to know what the rate of fading of interest for a Zynga game is, and I suspect one could rather easily figure this out using AppData.)

But the ultimate question in terms of maintaining high levels of overall user engagement is whether Zynga or any other company can continually pump out hit games.  The experience from other hit driven business models such as movies and video games, suggest that it is hard, and some companies can and some cannot.

So the questions in terms of overall users are how much room there is to increase the peak usage levels from the 1Q 2011 levels, how the inevitable fade in MAU levels can be managed, and how well Zynga can continue to churn out hits and refresh those usage levels.

This is one category of risks in the Zynga business model.

Zynga S-1: Zynga’s Golden Rule: Keep Mark Zuckerberg Smiling

Presumably, the 30% from Facebook Credits will keep Zuck smiling.  But, if not, Zynga’s fortunes peak and valley based on Facebook’s policies towards it.  The S-1 says that the Zynga business would be harmed if:

  • Facebook discontinues or limits access to its platform by us and other game developers;
  • Facebook terminates or does not renew our addendum;
  • Facebook modifies its terms of service or other policies, including fees charged to, or other restrictions on, us or other application developers, or Facebook changes how the personal information of its users is made available to application developers on the Facebook platform or shared by users;
  • Facebook establishes more favorable relationships with one or more of our competitors; or
  • Facebook develops its own competitive offerings.
Further Zynga says that its users declined after prior changes in Facebook policy:

We have benefited from Facebook’s strong brand recognition and large user base. If Facebook loses its market position or otherwise falls out of favor with Internet users, we would need to identify alternative channels for marketing, promoting and distributing our games, which would consume substantial resources and may not be effective. In addition, Facebook has broad discretion to change its terms of service and other policies with respect to us and other developers, and those changes may be unfavorable to us…. Facebook may also change its fee structure, add fees associated with access to and use of the Facebook platform, change how the personal information of its users is made available to application developers on the Facebook platform or restrict how Facebook users can share information with friends on their platform. Beginning in early 2010, Facebook changed its policies for application developers regarding use of its communication channels. These changes limited the level of communication among users about applications on the Facebook platform. As a result, the number of our players on Facebook declined. Any such changes in the future could significantly alter how players experience our games or interact within our games, which may harm our business.

Zynga is an asymmetrical relationship.  Clearly Facebook benefits by having Zynga generating activity and stickiness for Facebook, increasing user engagement and presenting more opportunities to sell display advertising.  Moreover, it benefits by the sale of Credits. But Facebook has all the leverage in the relationship, since Zynga cannot play off Facebook with another effective social network.  In that vein, Fred Wilson’s post today about rooting for the success of Google+ is spot-on:

My line about “don’t be a xyz bitch” is all about controlling your own destiny. These social platforms are awesome to build and launch on. They give you instant distribution, instant users, instant social identity. But in a perfect world you don’t want to be dependent on any single one of them. The more social platforms of scale there are, and we have a bunch now, including Twitter, Tumblr, and Foursquare, the better world it will be for developers. And our business at USV is investing in and helping developers build companies. So I’m rooting for Google+. I think it will serve users who aren’t being served well (or at all) on the social web right now. And I think it will be a strong new platform for developers. And both of those are great things for the web, our business, and entrepreneurs.



Zynga S-1: Facebook Credits (Platinum Status)

One of the questions that I posed about the Zynga S-1 filing concerned Facebook Credits.  I noted that the Zynga S-1 was likely to open a window into the size of the Credits revenue stream before the Facebook S-1, whenever that is.  Critically, according to the S-1, Zynga is paying the full 30% — i.e, list price and no super shopper discount for the Credits.  I haven’t seen any disclosure yet about whether there is some giveback by Facebook in advertising credits, so I assume there was not, unless there was no duty to disclose it which suggests it was not material (and it appears that Sales and Marketing Costs where advertising is accounted for went up slightly disproportionately last quarter which might suggest that any giveback is not material).
So how big is the Credits revenue stream from Zynga?  After the Facebook Credits cut, Zynga had approximately $222.5 million in “Online Game” revenue (i.e., purchase of virtual goods).  Using that figure and accounting for growth, let’s estimate that Zynga has a run rate of $1 billion in “Online Game” revenue this year.  Further assume that 90% or $900 million comes from Facebook.  This means that Facebook Credits will be collecting $386 million from Zynga for 2011.
Pretty impressive both in terms of revenue numbers (presumably pure income other than transaction processing fees) and the fact that Facebook didn’t blink and held to its 30% revenue share for the Credits, even for one its most substantial partners.

Zynga S1: Virtual Goods Paid With Virtual Money

The Zynga S1 is reported to hit the wire tomorrow, and certainly will reveal details about a compelling and, by all accounts, profitable business model.

One issue that I hope to see disclosure about is Facebook Credits — the currency that Facebook is requiring app developers to implement — that sends a 30% revenue share to Facebook.  Some potential questions that I look forward to possibly seeing answered are below, both about Zynga and perhaps even more interesting in starting to think about the Facebook revenue model and its down-the-road IPO .

1) Zynga is probably not paying 30 in revenue share, perhaps through a mechanism by which Facebook refunds Zynga some money back in advertising credits.  It will be interesting to see how that standoff settled out.

2) The result of that negotiation will reveal some information regarding the negotiating power that Facebook has to take the 30% revenue share, at least vis a vis its bigger partners.

3) We will start to get some data on how big the Credits business is for Facebook, as a big proportion is essentially Zynga revenues multiplied by the percentage that Zynga is paying for its Facebook credits.

4) While selling virtual goods is surely highly profitable, what is the effect of the Credits revenue share on Zynga’s profit model?  Is the business model so profitable that it can absorb a 30% cut?

5) Did the implementation of the Credits payment system in 2010 slow down Zynga’s growth, in terms of usage (perhaps because the increased costs were pass on to end customers in some way)?