The mental models of Chinese tech

On the reflexivity of fengkous

This newsletter took longer than I anticipated to write, but better late than never. I’m writing this edition from a quarantine room in China, just like the very first edition of the newsletter. Life has a strange rhythm, doesn’t it? Anyway, It’s good to see you again. 

For the past few months, I’ve gone native amongst the public market analysts.

I started by replacing the Patagonia, Arc'teryx and Common Project with premium mediocre business normcore but kept my Oura ring. Some status signifiers are too ingrained to yield. Once there, I followed the locals’ lead. I dropped ‘crushing it’ and ‘dude’ from my vernacular but stuck with ‘unit economics’ and ‘TAM’. One can’t be seen to be too familiar.

Progress has been steady. I ingrate myself by name-dropping W. Brian Arthur and Mauboussin to assure others of my intellectual credentials. My banal rants around valuations and inflation went over well. I’m hopeful there’s brokered common ground around Matt Levine. It is too soon to mention memes. Perhaps it will always be.

There have been close calls. One careless question left me cornered in a conversation about asset turnover ratios. I was on the verge of blowing my cover. After ruling out the possibility of fleeing while shouting ‘I like stonks’, I decided to take the tried-and-tested British route of ambiguous passive aggressiveness. “That is fascinating,” I replied, looking my challenger in the eye. But I couldn’t hold the stare. I lost the fight but lived to see another day.

Amongst my new tribe, I noticed a consistent fascination with mental models. The public market analysts revel in this strange delicacy. Mental models pepper casual conversations along the hallway, creep up in investment memos and take root as soon as they are uncovered.

It got me to reflect on a few things¹. Firstly, how mental models are used; secondly, the types of mental models that shape Chinese tech’s reality; and thirdly, how this came to be.

Used one way, mental models are algorithms that reveal patterns and insights in the world. Like algorithms, the refinement and evolution of models happen when they are run on new batches of data. We stop using models when they prove to be inadequate maps for the territories we’re navigating. Levelling up mental-models-as-algorithms requires knowing how to systematically find new data sources and adjust and adopt models to suit one’s purpose.

Used another way, some mental models are narrative framings that define and amplify a certain set of outcomes. Once mental models are adopted as our perception filters, they implicitly become guiding forces for our actions. Models move from being descriptive to prescriptive and accelerate various path dependencies. Call it reflexivity or viruses of the mind; once mental models take hold, they act as agents unto themselves.

What a man sees depends both upon what he looks at and also upon what his previous visual-conceptual experience has taught him to see. In the absence of such training there can only be, in William James’s phrase, “a bloomin’ buzzin’ confusion.” - Thomas Kuhn, Revolutions As Changes Of World-View

With its unique starting conditions, the Chinese tech ecosystem is a feast for the student of mental models. As I wrote in the Annual reflections of 2020, China is a land of scarcity mindset relative to Silicon Valley’s abundance world view. Most mental models used in the harsh mirror world are extreme variants of what we’ve seen in the West. Then there are those that have captivated the ecosystem so entirely that they’ve shaped an era.

In mid-2015, a series of mental modes gripped the Chinese ecosystem. The Fengkou era had a collection of overlapping models that formed FOMO like no other. They involved the following:

Fengkou / Follow the tailwinds - Success overwhelmingly derives from being on the right side of tailwinds. Best encapsulated by Lei Jun (founder of Xiaomi)’s witticism ‘A pig could fly if it finds itself in front of a fengkou.’ There’s no such thing as being the right person, only about being at the right place at the right time. This reframes opportunities as solely arising from market timing rather than from a visionary individual.

The Matthew effect / Power law - The market leader becomes more powerful over time while the rest of the pack becomes weaker. Popularised as the power law in the rest of the world where the market leaders get a disproportionate part of the value. 

For to everyone who has will more be given, and he will have abundance; but from him who has not, even what he has will be taken away. — Matthew 25:29

It is especially potent in high network effects areas such as consumer internet and therefore, places a special premium on being the first to market.

Owning the user by fulfilling their every need - With a defined market, growth arises from captive users and cross-selling them with a full suite of services. Capturing the user starts by owning their attention (through entertainment, status and utility) and bringing them down the purchasing funnel. This mentality drove the creation of the superapps and manifest in creating stickiness through community, content and commerce.

Reaching the ceiling / invisible asymptotes - There is a second inflection in the S-curve where a company hit the proverbial ceiling for growth and enter the business's maturity stage. Typically invoked by companies needing to find new growth drivers, it was first introduced by Eugene Wei in the post of the same name.

The interlocking of these mental models had momentous implications. Fengkous means that simply showing up to an opportunity is half the battle. The Matthew effect means one should be as early to an opportunity as possible. To avoid ceilings to growth, one should grow by owning the user and upselling them on various services. These foreshadows the battle royales in Chinese tech for years to come.

A critical mass must exist before a narrative frame can manifest as reality, though traction becomes inevitable as the mind virus spread from one person to the next. Eventually, reflexivity kicks in when self-appointed kingmakers such as VCs start exploiting these structures. The positive feedback loops feed upon themselves.

The first-order effect was that entrepreneurs and VCs tried to get involved in every fengkou as early as possible. The second-order effect became VCs actively encouraging and funding founders to join the latest fengkou. As more people started companies in a certain lane, the more fengkou-like a lane became, which in turn attracted more funding and easy growth for the early players. Early investors notched a healthy investment markup and took it to their LPs for their next fund. The entrepreneurs got money to burn for months.²

When the fengkou model became legible to Chinese entrepreneurs, the nascent ecosystem transformed into one that was marked by successive fengkous. Every single participant in Chinese tech thought about the current fengkou, argued over the next fengkou, and made career decisions based on where they thought the current fengkou was going. In other words, the legibility invoked reflexivity like nothing else and tipped the market into pure momentum.

With momentum investors up and down the food chain, the market will not catch up to the value investor. ‘Smart’ value-driven investments get passed on for their next rounds, and their investors are sidelined for not generating returns. In investment, the first place goes to being non-consensus and right, the second place goes to being consensus and right, there is no place for being non-consensus and wrong. Being a value investor in a momentum market is unwise, better mad with the crowd than sane all alone.³

Today, decaying bike remains half-buried in the side streets of China. Memento moris from the Online-to-Offline wars. The fengkou days are somewhat behind us. Instead, I am fascinated by the transmission of a new mental model for the years ahead.

Neijuan / Red queen world - In Chinese, the word means involution (内眷) - the opposite version of evolution. As growth dwindles and outcomes become bounded, competition increases for all participants as more effort is required to maintain the same output level. In other words, in the red queen world, one runs faster and faster to stay still. An example is when students are graded on a curve, everyone can study more but the results will fall along a distribution.

Where this mental model will take Chinese tech is still uncertain, but it is still as familiar as a zero-sum choice. Viewed one way, neijuan reduces effort as much as fengkou amplified it. When exertion of effort is framed as ultimately futile, what’s the purpose of trying? Viewed another way, everyone’s stuck on the same race together, and there’s no choice but to keep picking up the pace. To do or not do. How viciously does one want to fight for a bigger slice of a fixed pie?

With mental models, the maps become our territory. We create them and then they rule us. But what came first - Chinese tech’s zero-sum mental models or its scarcity mindset? The conclusion seems inevitable standing where we are today, but maybe another world was possible before the mental models took over. Moloch grows one transmission at a time.

Postscript

It is hard to see the nature of things in transcience. But it never feels like that until you are still. For the better part of 6 months this year, I’ve been living out of a suitcase. Traversing timezones, borders and sensible sleeping schedules. Now, I sit silently in quarantine writing this newsletter’s second chapter, wrestling with greater questions than ever before.

We spent 2020 and 2021 understanding aspects of Chinese tech as the landscape changed under our feet. For 2022 and onwards, the task at hand is to see where it is going. 

The narratives around China are as vicious as they’ve ever been, but how much of that are unexamined mental models filtering a set of reality back to us? I’m keen to go outside and see for myself. 

1. Yeah, we’re getting back into Chinese Characteristics wordcelling, don’t worry

2. In my mind, what differentiates this behaviour from the Gartner Hype cycle phenomenon is the level of meta-cognition players have about the game they are playing, hype cycle on Nash equilibrium as it were.

3. This depends on your time horizon, I’m modelling most career VCs in most single-stage funds as being unable to keep their seats after three years of non-performance. If one can ride the entire wave of the momentum market, all the power to them.