Pinduoduo: China Tech Series #1
Hey Folks 👋,
Earlier, I wrote a piece titled “China Tech vs. the World”. As I’ve come up the curve on the subject, I’ve decided to do a series of follow-on articles which probe China’s tech giants further. In this article, I look at Pinduoduo and intend to cover Meituan and BilliBilli in a few days. While deep-dives, these will largely be exploratory pieces rather than full-blown investment theses.
Pinduoduo’s numbers are as impressive as they come. It’s perhaps the brashest and most ruthless of China’s tech titans. The firm is a mere 6 years old, and as an e-commerce platform stands second only to Alibaba in China. It’s 2020 GMV of $255Bn stands at half of Amazon already. The “BAT” acronym was popularized in reference to Baidu, Alibab and Tencent as China’s answer to “FANG”. The eponymous Baidu has since fallen from grace and Pinduoduo and Meituan (Meituan leads PDD substantially in market cap) stand as contenders to that throne. While there are concerns about aggressive revenue recognition and governance (the firm does not have a CFO), there is no doubt on the impact PDD has had on China’s e-commerce landscape.
PDD founder Colin Huang is a fairly secretive individual. Per the FT; this secretiveness permeates the company’s culture where employees often refer to each by nicknames and are rarely aware of their colleagues’ actual names. Like Jack Ma, Colin is a Hangzhou native. Unlike Ma, however, he seem to have been a prodigy at the outset. In his youth he participated in math olympiads, and graduated from Zhejiang University - a member of China’s elite C9 League, the Ivy equivalent of China.
He landed his first job with Google, and was part of the team tasked with expanding Google in China. He quite after a three-year stint, and became a serial entrepreneur; starting and and selling a wide variety of online ventures. He counts BKK Founder Duan Yongping (of discount smartphone fame - BKK is the company that houses brands Oppo, Vivo and Oneplus) as one of his professional mentors. Recently, he handed over PDD’s reins to his consigliere and CTO Lei Chen, while still retaining control over most of the stock. Speculation abounds on the reasons behind this move.
Even by tech standards, the company’s expansion has been nothing of astounding. Its user base overtook JD.com within the third year of its founding.
This growth has come at the expense of profits – the firm has been in the red since its founding, and financial leverage runs in the range of 2.5x. It also seems to have been at the expense of the human element – a string of deaths at the company (one suicide, and one ostensibly related to overwork) has spurred debate about the infamous 996 work culture (9am to 9pm - 6 days a week but often longer) in China’s tech industry. Its business model and ruthless expansion has also displaced smaller retailers inviting regulatory scrutiny.
There has been much ballyhoo about “China Inc.” being a copycat phenomenon. My personal observation is that Chinese tech players, in particular, can boast of superior commercial acumen and more monetizable business models. A good point of comparison is Groupon – once a VC darling it has since faded into obscurity whilst its Chinese “copycat” moulded itself into the behemoth that is Meituan-Dianping. And this is actually a recurring theme in the Chinese technology landscape - WeChat’s utility, monetization and use-cases far eclipse Whatsapp; Alibaba leveraged it’s e-commerce business to redefine the fintech and payments landscape globally. Coming back to Pinduoduo, I do believe its business model innovations hold lessons for the global tech industry at large.
So what makes PDD so special?
They are pioneers of the so-call Community Group Buying phenomenon
Superior use of Data and AI. I’m not an AI expert, but they are are likely one of the best companies globally at leveraging data
Let’s look at these in turn…
PDD & Community Group Buying
The group-buying phenomenon is all about cost savings on bulk purchases. On the PDD App, users can form shopping teams to buy in bulk – or let a community champion (who handles the distribution of goods) do that for them. Bulk buying incentivizes producers to cut prices and eliminate middle-men distributors. High unlikely this works in the west, but this is tailor made for a country like China with abundant small-scale entrepreneurs and cost-conscious consumers.
Here is an overview of how the process works. WeChat is critical here for connecting group buyers.
Customers are able to aggregate orders and subsequently reduce the cost of last-mile delivery for suppliers. These savings are passed onto the customer. Around 55% of PDD’s volumes come from lower tier (sub tier 2) cities and this is very much phenomenon and platform for lower income consumers.
As you can see PDD is very much an offering for cost conscious consumers. This also accounts for the fact that while PDD’s MAU numbers are comparable to Alibab, the Gross Merchandise value is not. The goods sold are overwhelmingly of lower value. The image of PDD as a low cost app is something it has tried to shake off recently by offering Tesla Model 3 cars via its platform for instance. Group buying has nonetheless re-invented the supply chain in several ways. It has empowered small scale producers at the expense of retailers and middle-men.
Here is a slide about how PDD helps skirt the middleman. You have to imagine a small to medium scale agriculture farmer reliant on multiple middle-men. The reliance on multiple distributors is inevitable given the size and expanse of China’s geography. PDD steps in however, and effectively serves as an “Uber for Farmers”. Farmers have demand predictability and are also able to coordinate with end consumers and community champions to directly deliver produce.
The “Uber for farmers” connectivity is personally what I found most interesting. Agriculture being ripe for digitization in many developing countries; adaptations of the PDD model could potentially prove transformative as a model that than can be lifted and shifted to agricultural intensive and population heavy countries (India, Pakistan, Indonesia et al).
While PDD sells all manner of product, food and groceries are a key constituency. Around 12 million farmers sell their produce through the Pinduoduo. The company has start breaking out agriculture GMV separately in investor disclosures (still only at ~15% of total GMV though) - and is increasingly touting itself as an agri-business. This was in fact one of it’s counter-cyclical moats during the pandemic.
PDD’s Data Edge
Group-buying isn’t the only thing unique about the PDD business model. The data moats available to PDD are far richer than those available to conventional e-commerce players. One way they have cultivated this is via gamification and interactivity. PDD is quite interactive and tries very hard provide users with an entertaining experience.
On one hand, this is just to pepper over the dull experience of online shopping where one often mechanically searches for the product they need. On the other hand, this gives the company access to a wealth of additional data. A good example is the popular “Duo Duo Orchard” game where you water a plant through “droplets” acquired from various activities such as sharing links with friends. These activities let the company accumulate social data. For instance, people you play the Duo Duo game will likely be your close friends – so the game ends up providing social network data.
Data from group buying and social interactions are blended together to provide a highly personalized shopping experience and highly targeted advertising. It’s worth noting that around about 80% of PDD’s revenue come from advertisements. Merchants pay for visibility on the platform by bidding on keywords. And PDD’s placement mechanism is highly effective being closely tailored for each consumer. Effectively you have the precision and micro targeting of a Facebook like platform with the logistics prowess of an Amazon.
The chart of the left shows the additional data edge PDD has in terms of interpreting users’ context. Review Data, Gamification data, and and team purchase data supplement traditional behavior data and power PDD’s neural networks and“Distributed AI” infrastructure.
That’s all I had on Pinduoduo. Many thanks for reading! 🙏
