It’s hard not to dislike marketing for its manipulativeness. But, it’s also hard not to admire it for the insights into human behavior it brings.
This past week I’ve been immersed in the digital marketing space, attending three different events on the topic and talking with a variety of stakeholders, service providers, and personalities about what they’re seeing. Here are my takeaways.
Bottom line: We’re still in the early days of digital marketing and retail. Even though China is a first mover in the consumer space, everyone is still trying to figure out how to adapt traditional marketing and sales methods to the digital age. Technology continues to democratize and decentralize the creation and consumption of content, products, and services. Traditional companies, centralized by nature, need to deal with the rapid rise in China of new platforms, mediums, and celebrities.
On top of that, data management—how data is acquired and how insights are mined—is still a surprisingly large pain point for many MNCs in China. Lack of data is increasingly becoming less of an issue, while what to do with it and how to react quickly while a campaign is ongoing are the main challenges. However, while we’re all busy being amazed by the changes and powerful tools available, we’re still not asking the right questions about data privacy and the soon-to-be omniscient, omnipresent, and (perhaps) omnipotent power of AI.
Fragmentation
A brief timeline of major platforms for digital marketing
- Aug 2009: Sina launches Sina Weibo (now just Weibo).
- Jan 2011: Tencent launches WeChat.
- June 2013: Xiaohongshu (aka Little Red Book and Red) is founded by Miranda Qu and Charlwin Mao. Originally a social network for sharing travel tips, it has since become a hub for influencers and influencer marketing.
- 2016: Yixia Technology launches Yizhibo, a live-streaming platform, on Weibo. One of the leading players in the space, Weibo bought the company in Oct 2018. Yixia also operates Miaopai, a short video app used by many influencers.
- Sept 2016: Bytedance launches Douyin, a short video platform.
Content democratization = platform fragmentation: If social media is the canvas, then content is the brushstrokes. Sina won the Weibo wars (against Sohu and Tencent) in part because they were able to attract more influencers. At first, these influencers were media professionals, but the ecosystem quickly attracted movie stars, famous singers, and other high-profile personalities. With more famous people came more users. With more users came brands. With more brands came what we now call key opinion leaders (KOLs): those web celebrities making money purely through their online content. With China speed, of course, Weibo has already been displaced as the primary platform for digital marketing and not just by WeChat.
Here are a few smaller but still very influential platforms:
- Douban: Community-based website similar to Reddit where users discuss books, movies, music, and events.
- Keep: Social sports platform that offers personalized courses from a variety of coaches and personalities.
- Zhihu: Quora-like platform where users ask and answer questions on a variety of topics.
- Bilibili: Video-sharing site focusing on anime, comics, and games. A pioneer of the “bullet screen,” where user comments are directly displayed on top of the video, flowing from right to left, sometimes covering up the actual content.
- Tangdou: Social platform for sharing content related to “square dancing,” amateur dance meetups in public squares and parks, that mainly attracts middle-aged and elderly women.
- middle aged women. Users share tips and tutorials
- Chuman: Social network designed for minors interested in anime and manga idols. Users can make their own comics and share with others and also create custom avatars.
- Meiyou: Originally created as an app to track periods and female fertility; has expanded to include social sharing features.
While most online transactions take place on WeChat, Taobao/Tmall, and other e-commerce platforms, much of the marketing and messaging is done elsewhere. For brands, it’s quite a complicated space, as you can see, but many are learning quickly what to outsource and what to keep in-house.
The power of WeChat: Since its launch in 2011, WeChat has become the primary destination for China’s mobile internet users. However, with the rise of more and more niche platforms and Bytedance’s extremely compelling products, the super app actually saw a dip of 8.4% in average time spent in the app from Dec 2018 to June 2019. But there is a silver lining: average time spent in mini programs (WeChat’s version of the instant app) is up 23% and monthly active users have reached 74 million. For the platform and for brands, this is great news.
Launched in 2017, mini programs took some time to get off the ground, but now are proving to be amazingly effective at creating entry points and engagement channels for brand retailers. After payment, WeChat users are prompted to open the mini program to join the membership program where they can receive personalized recommendations and special offers. Brands can even open their own store on WeChat through a mini program whether that’s through partners like JD.com or developed in house, giving them increased access, engagement, and sales.
Deals with divas: KOLs have been a powerful marketing device since at least the 1950s with the advent of the Tupperware party. With Weibo, KOL marketing really started to take off, but fast forward 10 years from Weibo’s founding and it’s still an immature space. Brands and KOLs are still trying to find the best ways to work together so both can make money, but their interests don’t always align. Both are also constantly challenged by the rise of new platforms, as well as new rules and regulations from both government and the platforms themselves.
Case in point: Xiaohongshu was taken off both Android and iPhone app stores after they failed to resolve fake reviews, artificially boosted page views and purchase numbers, as well as tolerating ads disguised as product reviews. This was after the platform implemented strict new rules that reduced the number of KOLs allowed to operate on the platform from 17,000 to 4,500.
KOLs in some product categories like health and beauty are even competing for brands themselves. As competition increases and individual KOLs see their margins declining, many have branched out into creating their own brands. Many OEMs, struggling with overcapacity, are more than happy to work with a wide range of smaller brands to offload their products.
Data
Data overload: Online or offline, data gathering has been a challenge for brands in China. Offline, the retail supply chain is still very fragmented and, especially in fast-moving consumer goods (FMCG), it can be difficult to track sales. Online, actually figuring out where the data was and creating effective pipelines was difficult. However, many brands have resolved these problems by establishing in-house data collection mechanisms. Combined with second-party data (data shared between two or more parties privately) and third-party data (data bought from outside sources), brands have built powerful and sophisticated dashboards that show real-time engagement and sales performance. However, now the struggle is to move decision-making (in targeting, pricing, and even production) from humans to computers.
Where this is all going
Prepare to be manipulated: AI, as I’ve written previously and as most of you already know, has enormous potential to change the very fabric of economics, politics, and society. However, it is still not sophisticated enough. Sure, you can get “personalized” recommendations based on past purchases and behavior, but dashboards I outlined above are, in the parlance, limited to diagnostic analytics: with the right tools and expertise, data can be mined to understand why a campaign was or was not effective. However, the real power of AI comes in prediction (understanding what will happen) and prescription (making something happen).
Like it or not, marketing is manipulation. Whether that’s for the common good, as Cass Sunstein suggests, or for private gain, as in Google’s experiment with Pokemon Go. And, indeed, this is the end game for AI developers: profiling and segmenting each and every one of us so we behave in the ways according to the plans developed in board rooms and government meeting halls.
The data privacy argument, in this regard, is completely misguided. It’s not about whether or not Alibaba or Apple is listening to your conversations with their AI assistants (they actually have to have a human monitor the machine to make sure the models are correct). The data privacy debate needs to shift to regulating how data is collected in the first place and how it is being used. Right now, almost all human behavior is the raw material, as abundant as the air we breath and expanding every day.