Thoughts on Status as a Service, utility as the ultimate moat for social networks and how the framework breaks down upon application to videogames

Eugene Wei’s post on understanding social networks as “Status as a Service” businesses was brilliant. It was full of insights and new mental models for investing in Facebook, Snapchat, Tencent and Twitter. I’ve aggregated my two Twitter threads on “Status as a Service” into this post. I generally try to post original thinking/content on Twitter, Medium, etc. but Wei’s post sparked so many thoughts for me that I had to respond. In a strange way, I feel like I am back in college and writing a critique of an important book. I do disagree with his conclusions about videogames but am still thinking differently about videogames as a result of his post.

Wei’s conceptualization of social networks as ICOs with unique, graph specific “proof of work” helps explain why so many of Facebook’s attempts to copy features from other social networks failed yet stories succeeded. “It’s the unique combination of a feature and a specific graph that is any network’s most critical competitive advantage.” Specific features (proof of work) are social graph dependent in terms of their success with “features” being the “proof of work” required to earn social capital. This explains why so many of Facebook’s attempts at copying features failed: Facebook had a different graph (everyone you know) vs. other social networks. Stories for Facebook succeeded because they solve a universal issue for all social graphs: “supply side social modesty.” Other features Facebook tried to copy were graph dependent, hence the failures.

“Another existential risk that is somewhat unique to social networks is this: network effects are powerful, but ones which are social in nature have the unfortunate quality of being just as ferocious in reverse.” MySpace being a great example of how quickly network effects can unwind. His most powerful point for investors is that utility is the ultimate moat for any social network given that social networks can unravel quickly as described above. The inverse of Chris Dixon’s excellent post “come for the tool, stay for the network.” Riffing on this, I’d say “come for the tool, stay for the network, never leave because of the utility.” Build defensibility by making your network useful. WeChat succeeded at this — partially due to an easier road in China (lower credit card penetration) — and Facebook has failed to do this so far.

It is easy in the United States to eliminate Facebook from your life and impossible to eliminate WeChat from your life in China. As a sidenote, I highly recommend Connie Chan’s excellent post from 2015 on WeChat. Having visited China either annually or every other year, I will say it is difficult to function there without WeChat and becoming more difficult every year that goes by. Wei correctly posits that Facebook would be in a much better position today if it had persisted in trying to compete in search (Google’s main fear re Facebook), developed a payments system or developed an assistant (likely required success in search). Those would have all increased utility for Facebook and put in a stronger, more defensible position by making it harder to deactivate.

Instead he sees Facebook as having abandoned all of these efforts prematurely. Nothing could compete with advertising as a business opportunity but that is/was ultimately dependent on engagement, which is now a point of vulnerability for Facebook as a result of their failure to build utility. I will never forget Sheryl Sandberg’s response upon being asked about low hanging fruit (from a monetization perspective) at Facebook in 2014ish: “We haven’t even gotten to the low hanging fruit. We are still stepping on fruit.” Facebook’s engineering efforts were focused upon 1) monetization and 2) sustaining engagement by rapidly copying features (proof of work) from other social networks. Nothing had a higher ROI in the short term than the 1) and 2) seemed eminently logical. Being a fast follower on features was rational — it worked for Microsoft for years.

Yet all the efforts to drive engagement via added features failed with the exception of stories for the reasons described above. Social networks are not operating systems. Feature success is graph dependent. This also explains why Facebook’s efforts to unbundle their app (Groups) failed with the exception of Messenger. All of the utility of the features they were copying and unbundling were graph dependent and Facebook was using the same graph, thinking that was their advantage instead of realizing it was also their disadvantage.

Going back to earlier points, understanding that the success of “new forms of proof of work” is graph dependent is critical. I wonder how many of Facebook’s failed features would’ve succeeded if they had let users easily change the associated graph. The focus on copying features doomed to failure because their success was graph dependent led to the failure to develop any real utility — search, payments — which left Facebook, especially big blue, vulnerable.

The fact that utility is the ultimate moat is why users consistently say they would remove all Facebook apps before Google apps in surveys despite valuing Facebook at $1000-$2000 in other surveys. It is semi-impossible to function today in America without Google but easy without Facebook. Facebook seems to realize all of this now. Hence the deployment of several top tier executives to their cryptocurrency project, which could be a big unlock for payments and the utility of all their networks and thereby increase the sustainability of their engagement. And their go forward emphasis on sharing with smaller groups and adding more ephemerality, which could make it easier for them to succeed with features where they had previously failed. They are much more focused on utility than ever before.

Eugenue Wei identifies enforced graph portability as the critical regulatory vulnerability for Facebook (I would add being broken up although the constituent parts would likely go on to collectively perform very well Standard Oil style). The graph does have value. It just has to be correctly applied. And it’s not really relevant to any of my points, but his conception of the transition of the transition to an algorithmic feed as an interest rate hike that decreases liquidity on a network is incredibly smart.

However, I respectfully think Eugene Wei’s framework falters upon application to videogames, especially multiplayer games. As I’ve written before, videogame franchises are more durable than generally believed.

Wei posits that “Video games illuminate the proof of work cycle better than almost any category…the drosophila of this type of analysis given its rapid life cycle and overt skill-versus-reward tradeoffs. Why is it…that big hit games tend to have a life cycle of about 18 months?” This is incorrect — multiplayer games last longer than 18 months. Some of the “drosophilia” like World of Warcraft are as old as thefacebook.com and Call of Duty is one year older. I am going to try to explain how and why these games have been so durable, which confirm many of Eugene Wei’s recommendations for Facebook and other social networks.

“A new game offers a whole new set of levels and challenges, and players jump into the status competition with gusto. But, eventually, skill differentiation tends to sort the player base cleanly.” Skill based matchmaking (SBMM) in many multiplayer games obscures this “sorting.” “Players rise to the level of their mastery and plateau. Simultaneously, players become overly familiar with the game’s challenges; the dopamine hit of accomplishment dissipates.” Not in multiplayer games: players don’t rise or fall to their true level of mastery because of matchmaking.

Multiplayer videogames manage Player vs. Player “status as a service” via matchmaking. Even unskilled players get a dopamine hit every 3–7ish games as the matchmaking systems ensures they get a win or an improved/positive K/D ratio. Everyone thinks they are a better player than they actually are. This is why a large player base is such an advantage for a PvP centric multiplayer game — it is easier to efficiently administer dopamine hits in a variable way in-line with B.F. Skinner’s research and thereby manage “status as a service.” Skill based matchmaking is always being toggled. And a large player base makes the gaming experience in and of itself better — more exciting games, less lag, etc. While winning is a dopamine hit, winning a close game is much more powerful — a “peak” experience. Metcalfe’s law(ish) in videogames.

The ways videogames use matchmaking to administer “status as a service” is a roadmap for social networks in that they can lower the bar for new, unskilled players (boost engagement/likes) and raise the bar for more skilled players as Wei suggests. As a sidenote, most videogame co’s deny using SBMM (players *think* they dislike it — mainly because streamers *actually* do dislike it) and claim to only use connection based matchmaking. This is not the truth. They almost all use skill based matchmaking(along w/ connection based matchmaking) to varying degrees and with varying frequency. A famous streamer named Drift0r made a bet that there no one could find evidence of SBMM in Call of Duty and he had to pay out $1000 after a Reddit user (I believe) submitted fairly clear statistical proof of SBMM.

Multiplayer games are also actual social networks that I suspect follow the Dunbar number. A casual perusal of the size of the average friends list on Xbox Live shows a clustering around 150. I would love to see research on this. Steam, Discord, XboxLive and PSN are actually fulfilling the dream and promise of Path.

The combination of 1) managing status as a service through matchmaking and 2) being satisfying, Dunbar level social networks are why multiplayer games like Counterstrike: Global Offensive, League of Legends, World of Warcraft, Rainbow Six Siege, Call of Duty, etc. have been so durable.

However, I am thinking about videogames differently as a result of Wei’s post. “The danger of having a proof of work burden that doesn’t change is that eventually, everyone who wants to mine for that social currency will have done so, and most of it will be depleted…At that point, the amount of status-driven potential energy left in the social network flattens. If, at that inflection, the service hasn’t made headway in adding a lot of utility, the network can go stale.” Utility equals dopamine hits in the case of videogames. Humans like variety. “One way to combat this, which the largest social networks tend to do better than others, is add new forms of proof of work which effectively create a new reserve of potential social capital for users to chase.”

In videogame terms, as identified by Wei — new forms of “proof of work” are absolutely essential to keeping the game vibrant — not to enable new status games, but to keep the existing status games interesting. New maps, weapons, characters, seasons, modes etc. For single player games with a multi player game bolted on, this necessitated entirely new games — and allowed more upfront monetization — but the world is moving away from this model (Counterstrike: Global Offensive, League of Legends, Fortnite, Rainbow Six Siege, Apex Legends, Player Unknown Battlegrounds etc.).

Games like Battlefield that require one to start over every 1–2 years are like social networks that annually take your follower count, likes, etc. to zero. This is becoming an outdated way of managing status as a service for the industry. Finite games are becoming infinite games. This is why Destiny2 was so detrimental to the franchise. Players missed their Chatterwhite shaders, etc. All existing status was prematurely destroyed. And Destiny2 was much smaller than Destiny1 at the time of release — fewer Raids, strikes, etc. It is much better to continuously bolt new forms of proof of work on to an existing, always evolving game than start from zero every year. And monetize through those new, more incremental forms of “proof of work” over a steadily growing player base. Games as a service. = )

I would posit that videogame publishers should only take the risk of taking status to zero via a sequel when the incremental return on new forms of proof of work go down dramatically. This will be a transition but I do think there are ways for videogame co’s to manage this. Free to play Battle Royale modes for every shooter!

In all seriousness, I believe all core PvP elements of every game — including maps- will become free to play as that will be the only way to get a large enough player base for quality matchmaking. PvP will be monetized through characters, items, cosmetics, emotes, skins, etc. The portion of the player base that enjoys single player or PvE content can then be monetized in more traditional ways — players can pay upfront and/or annually for new PvE focused campaigns, content (horde modes for Battle Royale games). PvP thus becomes an acquisition channel. This is likely a more lucrative business model long term. It is much easier for players to accept paying for DLC when they did not pay $60 for the game. And we know from mobile games — and Fortnite — that free to play models can create larger games in terms of revenue than games where the players pay $60 upfront. The new business model will be the inverse of the old, which was anchored in a single player, PvE centric worldview with multiplayer PvP bolted on. The new business model will be anchored by a free to play multiplayer PvP experience with a PvE experience bolted on.

Finally, the Marshmello concert in Fortnite was important. The bias so many have against videogames is preventing smart people from seeing that a videogame might serve as the basis for the next version of the internet i.e. the Metaverse/Oasis.

--

--

--

Founder, CIO & Managing Partner, Atreides Management LP. Former Portfolio Manager, Fidelity OTC Fund. No investment advice, views his own. More: gavinbaker.net

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

Feeling uncomfortable DMing someone new on Twitter?

They’re Trying to Sow Division and Our Minds are Rich Soil

Robin Hood and easy access brokerage accounts are the new Milennial Crack

A Call to Action

How To Make Money on Fansly?

The Stanford alum behind Netflix’s “The Social Dilemma” wants you to stop scrolling

a sneak peek at the new twitter

AOL Instant Messenger’s Significance in digital history

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Gavin Baker

Gavin Baker

Founder, CIO & Managing Partner, Atreides Management LP. Former Portfolio Manager, Fidelity OTC Fund. No investment advice, views his own. More: gavinbaker.net

More from Medium

We can both care — we just don’t need to care in the same way.

The issue with digital identities

The ‘Parties’ Debate, 14 January 2022.

Biden Announces BUB to supply Baby formula to the United States: Fundamental Outlook