Operations: The Chinese Internet Management Innovation
This newsletter translates articles from Chinese media about tech, business, and political economy. I also host the ChinaEconTalk podcast, where this week I interviewed a Canadian researcher about the broader East Asian AI policy landscape, looking comparatively at China, South Korea, Japan and Canada.
Most leading Chinese tech firms’ org charts are copied directly from western tech giants. Yet in Silicon Valley, operations departments are seldom found. Operations, a mix of marketing, sales and customer service, has been crucial to the rise of Chinese tech giants. Perhaps the best American analogy is the all-important ‘ground game’ of presidential campaigns.
What follows is a full translation of an article exploring operations in Chinese tech. It argues that widely varying market conditions, cheap labor and poor data management lead companies to opt for operations as opposed to growth hacking.
Why Don’t American Internet Companies Have Operations Departments?
T什么美国互联网没有“运营”岗？Huang Youcan, August 6, 2018.
About two months ago, a PM who worked in a well-known Internet company in Silicon Valley returned to China and asked a mutual friend to introduce us. He had discovered that in China people everywhere seem to talk about “operations specialist” positions, which do not exist as such in American internet companies. Confused, he asked me to clarify what “operations” means in this context.
Considering that the US is the birthplace of the internet, it’s not surprising that many aspects of China’s internet—from business models to product concepts—imitate American practices. However, “operations specialist,” an extremely popular job title in China, accounts for about 3-4 times as many employees as product managers.
For a long time, the “operations” position has remained a vague concept. Even people with that job title have difficulty describing what is it that they actually do. So our first step is to reach a common understanding on what “operators” are.
In the book Methodology of Operations, the basic definition states: “Operations acts as a medium to help the product and consumer establish a better connection.” Reflecting specifically on what operators do, we find two main functions: 1) creating leads, gathering and converting traffic, and 2) user management and service.
This definition is generally accepted throughout the industry. In theory, “operations” is a problem that all internet companies need to face and solve at some point.
However, in comparison to the scope of work that China’s internet companies assign to “operations,” American internet companies do something completely different. Generally speaking, the same work is achieved through sales and advertising.
The question is why?
First off, China’s internet is much more complex than the US internet.
Specifically, social structure and class differences are relatively more stable and homogenous within the United States; geographical distances between cities are relatively small; and everyone’s basic needs and demands seem relatively standardized (for example, a group of American classmates is generally happy to order pizza together, disagreeing only when it comes to the toppings). This creates set ways for dealing with the vast majority of users’ needs.
However, in China the situation is different. China has a huge population and a very stratified society. User needs differ widely between first-, second-, third-, fourth- and fifth-tier cities, not to mention villages and smaller towns. Diverse age ranges, personal interests, and hobbies further result in a wider range of needs.
Moreover, in China, it’s not only the big markets that require multilayered segmentation. Even companies that have a specific group of users will frequently need to vary its services.
For instance, let’s take a look at Dianping [best described as Chinese Yelp + Uber Eats, but now much more]. Their users could be divided into the following several subgroups:
Users who simply enjoy eating out. They use Dianping to search for restaurants that meet certain conditions and have excellent reviews.
Users who only follow recommended places and lists (usually food bloggers or serious foodies). After visiting a restaurant, they can be relied upon to write an extensive review of the venue.
Users who fall somewhere in between the previous two. They will occasionally write a review.
Users who use Dianping as a reference. They read restaurant reviews and then order delivery via Meituan.
Every type of user operates under a different logic; therefore, every user group should be approached in a different manner. To go a step further in the case of Dianping, a platform heavily reliant on its customer-generated reviews, is not easy to figure out how to motivate each group of users to leave a useful review. A strategy to solve this issue could involve setting up a reward mechanism for each category of user: organizing events (free meals for super-users at their favorite restaurants), reengaging with inactive users, etc. Because these results are difficult to achieve through algorithms, there is a need to invest in manpower. Enter the operations specialist.
An operations job is focused largely on creating a specific atmosphere for users, but it’s not so directly connected to the final sale—which is what the sales team would focus on.
That said, there is no harm in doing a brief comparison of China’s Dianping and its American counterpart Yelp. Yelp’s homepage is simple. It doesn’t have a fashionable lifestyle area with a constant stream of new content. It doesn’t have multiple banners advertising events and videos, nor an option to follow creators. In short, Yelp doesn’t have any diversified or creative properties. Instead, it’s a tool for searching information about nearby restaurants.
The bottom half of Dianping’s homepage features a social feed users can scroll through to discover new restaurants and activities
With this type of product, creating an independent service department wouldn’t make any sense. To be clear, it’s not the case that Yelp does not engage in “operations” activity. From time to time, users receive emails from Yelp (in the US, email is the most popular way of promoting content), either to recommend nearby restaurants, ask about the user’s preferences, or remind us that we haven’t used their platform in a while. This type of email is usually implemented after testing through algorithms.
Most US internet companies probably work this way. In the United States, where engineering culture dominates, attracting and establishing a deep connection with users is seen as a problem to be solved by machines, not people.
Chinese internet companies operate on a simple and common belief: There is always a growing demand for a more personalized approach. Customers respond to a richer, more complex system; as they are confronted with more variables, more manpower is required to follow up with these variables, and subsequently the company cannot rely on a simple, standardized approach to satisfy the needs of a large number of users.
Thus, on the Chinese internet, many products are in need of continuous “service,” even if that means organizing a different activity each week or providing every type of user with personalized content.
Cheap manpower (compared to the US) and a surplus of low-end labor in China is another reason for the prevalence of the “operations” job title. Much of the growth in customer service relies upon large-scale manpower and a lot of simple, repetitive labor.
In the end, we may need some more time to discuss the most popular user growth model of the US internet, which relies on data-driven growth. The main question we will explore is: Why isn’t such a growth model just as mainstream in China?
If you follow tech news, you probably know that in the past 2-3 years, a strategy of “growth hacking” has become popular and gained support in China. The core of this idea is precisely the need to rely on data and technology to drive growth.
However, this method has only successfully been implemented in China in a few cases thus far—one cannot say it is a universal thing here.
The core reasons behind this are twofold:
The first reason is related to the rich internet user base and their diverse needs in China. The key to data-driven growth is finding a correlation between user behavior and growth. In business, the simpler that user behavior is, the more standardized an approach one can take—hence, it is easier to find such correlation.
The second reason is that in China, data analysts are extremely rare and highly paid. According to LinkedIn’s “2016 Chinese Internet Talents Report,” data analysts are ranked as the sixth-most sought-after position within China’s internet industry.
Thus, calculating from a perspective of costs and considering the salary of entry-level jobs, 5-6 lower-level operations specialists may equal the cost of a single data analyst/engineer.
For all of the aforementioned reasons, the majority of Chinese Internet companies apply and collect “data” on a very basic level. Even companies that received Series A financing may be unable to perform full core data monitoring and analysis. If both data monitoring and analysis are nonexistent, what is there to be done?
These are two main reasons why the Chinese internet industry relies on “operations, even though many people agree that “data-driven growth” is more scientific and efficient. Setting up the latter type of system would be difficult for the majority of internet companies. It’s simply easier to recruit a bunch of operators.
In short, we believe that the key reasons for the dearth of “operators” in the US internet industry are as follows:
China and the United States have huge differences in both social structure and user diversity—with Americans being comparatively more homogenous, while Chinese are extremely diversified.
China’s labor force is both large and relatively cheap, while the US has higher labor costs. This has led to companies in China relying on human resources to carry out many simple tasks related to user growth and maintenance, while in the US everyone is trying to solve emerging problems through technological and standardized processes.
Americans naturally respect the rules, whereas in China’s business world, everyone “naturally” gets accustomed to profiting by exploiting blurred lines and pushing the boundaries.
Among US Internet companies, “data-driven growth” is the most popular model, while in China it is not universally practiced because: a) Users’ diverse demands make it difficult for the majority of early-stage businesses with a non-standardized product to find a relevant model to help drive growth; and b) In the internet sector, qualified data analysts/engineers are scarce and costly, which is an obstacle to companies being able to fully exploit the value of data.
While in Bangkok this past weekend, I stumbled across The King Never Smiles. Written in 2006 by a longtime Thailand-based journalist, is a fascinating dive into Thailand’s modern political history. After spending so much time reading exclusively about China, it’s refreshing to have my mind stretched trying to understand the dynamics of a system starring Buddhist royalty. The king over fifty years evolves from afterthought to the center of Thai politics through a combination of image management, philanthropy, tactical domestic political alliances and American Cold War support.
The old queen’s birthday was coming up.