In life, when confronted with a mystery I always use this principle:
Occam’s conspiracy razor: Never ascribe to malice when stupidity will suffice.
In 2012, Zynga experienced a rapid decline in numbers while at the same time mobile gaming and mobile traffic took off. The expectation that mobile would take over wasn’t a surprise. We all knew it was going to happen, what was surprising was how fast web based browsing dropped.
Mobile was going to be big, but why was it at the price of web? Why did the web die out?
Occam’s conspiracy razor was that people just wanted to move to mobile, and that 2012 was the year it happened. There was no conspiracy, no external event, just a mass migratory movement.
In early 2012, as I looked over the landscape of mobile internet, my reaction was that this was going to be a platform that added more people, not took away from the web. Adults would both use the mobile device and the desktop to do all of their activities. Most people, sitting in front of computers during working hours, would play their games, do their Facebooking, do their shopping on their computers while working. Later in the day, when they went home they would use their mobile devices on the couch while they hung out with their families and friends.
That. Did. Not. Happen.
At the end of 2012, we could confidently call the death of the desktop web. Desktop DAU for Zynga cratered, and Facebook web DAU had flatlined.
What happened was that while Zynga and Facebook were creating the ultimate form of escapist fun for the masses, corporate America noticed.
And what they noticed was that web-surfing that had been an annoying tax on employee productivity was becoming a massive time sync. Employees were playing their Zynga games, sharing on Facebook with their family and friends, instead of doing theirs jobs and incidentally consuming vast amounts of bandwidth. Later Netflix entered the picture and the amount of lost productivity and bandwidth was getting serious.
Before Facebook, social non-professional interaction was hidden in corporate and private email or the phone. The water cooler, the smoke break, the lunch room, the break room where we hid minutes of wasted time every day.
Before Zynga, it was really hard to play a game on a corporate laptop. The security teams locked computers down so tight, that nothing could be installed. And to be honest, you’re not going to take a 5 minute break to play some Call of Duty. Most employees played Minesweeper because that was the only game they could install.
So lost productivity was visible but unblockable.
Enter Zynga. Farmville at it’s peak had 30 million DAU. That’s an insane number for a game. 30 million people were not playing at home. Peak DAU hours were during the working day. Corporate America noticed.
At the same time as Facebook and Zynga were taking over the world, ng-fw‘s (next generation firewalls) came into existence. Their claim to fame was that they could identify the applications and then apply security policies at the application layer.
Apparently, as I recently discovered, for the last several years the basic pitch of an ng-fw begins with:
How do you stop your employees from playing Farmville?
Obviously it’s more nuanced … Not really. I mean sales guys as recently as last week positioned ng-fw as how to block people from playing Farmville. The thing was that corporate america figured out how to control what applications their employees were using, even when they were on the web.
Managers wanted to stop their employees from goofing off, and since employees were using communication channels that the managers could choke, management did.
What about SSL? Well it turns out that the firewall vendors figured that out as well. The firewalls would terminate the SSL connection to the external site, and then in turn re-sign the certificate. The IT teams would helpfully disable the warnings associated with the re-signed certificate making people oblivious to the fact that their traffic was being man-in-the-middled.
Net effect, throughout corporate America, firewalls were quietly and silently blocking access to consumer web-sites that employers felt were not really related to work.
So what does this have to do with Zynga?
At the same time, at Zynga I was observing this really weird unexplained phenomenon.I was seeing evidence of people trying to start our games and failing. Stuck in the cauldron of Zynga, we had no insight into what was going on.
We assumed, applying Occam’s razor, that people were starting to load the game and then quitting. The sheer scale of attempts was mystifying, but we assumed that people had their reasons.
I mean, how else could games be blocked?
Last week I went to Hawaii and experienced ng-fw first hand.
1. Sometimes I couldn’t load the game.
2. Sometimes I could load the game but then the connection would be dropped.
And I realized that our customers had been blocked from accessing our games.
And then it clicked. All of those mysterious reports, all of those users who couldn’t get to our games, all of those bizarre late night sessions trying to understand what the hell was going on, and we missed the most obvious thing of all:
Employers didn’t think their employees should be playing games at work and were doing something about it.
So why was this bad for Zynga?
Let me caveat the next bit with the following: Mark Pincus never accepted the idea that Zynga was a victim. We didn’t fail because other people did things to us, we failed because we didn’t execute because we didn’t deliver. And I agree. If we failed, it’s because we failed not because other people screwed us.
Let me also caveat, that this is pure speculation. I don’t have any data to back any of this up. At the time I didn’t know what to collect, and now I don’t have access.
DAU and Engagement
Spend enough time on the Facebook platform, and you know that the only thing that matters is engagement. Facebook wants people to be engaged with their product, they’ll flood you with users to see if they’ll stick, but if you don’t keep those users be afraid, because eventually they’ll point those users somewhere else.
Firewalls were breaking Zynga’s engagement in two ways.
The first was that users were unable to get to the game. With games that had a plant-harvest cycle, if you can’t make it into the game to harvest, then you’re likely to quickly give up with the game itself.
The second was that users were conditioned to stop clicking on our feeds and virals. If clicking on the feed or viral would result in nothing because of firewalls, people stopped clicking on them.
As a result Facebook was seeing a decline in engagement in Zynga games. And because they care about their users, first and foremost, they started to point their users away from our games. And good for them.
Unfortunately this fed a vicious cycle of decline. The more our users left, the more we tried to reach them through virals, the more Facebook users got annoyed with those virals, the more Facebook throttled our virals. And because Facebook was always looking for some other content they could get people engaged with, Facebook pointed the firehose of users elsewhere.
If there was no mobile platform, I posit Facebook would have seen a flatlining of users as did Zynga. Instead what happened was that people discovered the one device that had a network their corporate bosses could not control, their mobile phone.
With their access to Facebook blocked, employees discovered what teenagers have known since forever, mobile phones are the only way to talk to your friends without anyone getting in the way.
Much like teenagers who used SMS as a cheap way to talk forever with their friends, adults reached out to the mobile device as a cheap way to keep connected once their internet access got blocked.
And mobile Facebook traffic skyrocketed.
And once everyone got used to using their mobile device instead of their corporate laptop to connect with the Internet, the rest was history.
The transition was inevitable, and firewalls forced the transition to happen ridiculously fast in 2012.
Some final thoughts
When I look back at my own mistakes, what I missed about mobile was not the form factor but the access to the users.
At the end of the day, consumer websites live and die by their ability to reach end-users. And unfortunately, the last mile is controlled by corporate entities that may have a dim view of people goofing off on the job.
Mobile is the future because it represents the only reliable way for consumer businesses, net-neutrality excepting, to reach their customers without some intermediary blocking them.
What I missed at Zynga wasn’t mobile, it was the fact that only on the mobile platform could we be guaranteed access to our users.