PDD

The Genius Hiding Behind the Gimmick

After digging into PDD (Pinduoduo’s parent company), I’ve come away thinking the business operates in a way that’s almost ridiculous, except the ridiculousness is the genius. What looked at first glance like gimmicks turned out, on closer inspection, to be some of the smartest psychological engineering I’ve seen in e-commerce.

Take the “gamified shopping experience” framing. On its face, that sounds like a marketing buzzword, maybe even a little embarrassing. But any investor who’s spent real time in the market knows that psychology is the game. Staying emotionally stable, resisting FOMO, not panic-selling when a position drops 30% overnight, these are far harder than they sound, and most people fail at them repeatedly. If you’ve sold anything, in my case medical supplies, you already know that people buy from those they trust and like, and that scarcity and pressure move behavior in ways logic alone never does. Any business that ignores these levers isn’t trying very hard.

PDD built its entire model around them.

The Group-Buy Origins

PDD’s original model was extremely simple: group-buy fruit orders. On the surface, it’s just a bulk-discount mechanism. But look at it as a game, and it’s brilliant. The more people you recruit into your order, the cheaper the price gets, so customers start doing PDD’s marketing for free. You send the link to your neighbor, your grandma, a dozen contacts because you want a cheaper price on fruit.

This parallels Adam Smith. In The Wealth of Nations, Smith’s insight was that capitalism works by channeling individual self-interest (greed, if you want to be blunt about it) into outcomes that benefit society. PDD does something structurally identical, just at the level of a single company: it channels customer frugality into free marketing and free distribution. Customers aren’t trying to help PDD, they’re trying to save money, but the byproduct is that PDD gets an army of unpaid salespeople and distributors, all self-organized, all cost-free.

Manufacturing Daily Habit

The second layer is attention capture, and here PDD borrows straight from the Google/Facebook/TikTok playbook: build a habit loop that gets someone to open the app every single day, whether or not they intend to buy anything.

The clearest example is the virtual fruit tree. You log in, water a cartoon tree. Do it enough days in a row, and PDD ships you actual, real fruit. The mechanic itself is almost silly, you’re rewarded for watering a picture, but the function is completely serious: it trains a daily reflex. Just like your thumb automatically opens Instagram or TikTok before you even consciously decide to, PDD is training that same involuntary muscle memory, except pointed at a shopping app instead of a content feed.

PDD in a sense is really selling habitual attention.

The Margin Tell

Once you see PDD as an attention machine, the margins start to make sense, and they’re the first thing that should draw your attention.

Retail, as a category, runs on brutally thin margins. Look at JD.com, a genuinely well-run business with excellent fulfillment and logistics. But that kind of infrastructure is extremely capital intensive, so it’s no surprise JD nets somewhere in the 1% to 3% range on the products it sells. PDD posts net margins north of 20%, close to ten times what JD earns. Two companies moving the same broad category of goods don’t end up ten times apart on margin unless one of them is doing something fundamentally different underneath.

Traditional e-commerce is not where PDD makes its money. In the domestic business, the platform takes a small cut on each transaction, but the real engine has always been the attention it captures and resells: advertising and marketing placements, flash-sale slots, promotional positioning that merchants pay up for. PDD doesn’t hold inventory, doesn’t buy product, and doesn’t own logistics centers. It sits on top of manufacturers and consumers as an interface, charging for visibility rather than for the goods that pass through it.

Because the reported statements have shifted. On a consolidated basis, transaction services revenue has now grown to rival and slightly exceed marketing revenue, but that’s Temu’s doing, since Temu’s marketplace fees book into the transaction services line. Strip Temu out and the domestic picture is the one I just described, a business that earns far more from selling attention than from taking a cut of transactions.

The Capital-Light Thesis

This is a double-edged sword, but I like the model. The race to build “the best distribution infrastructure” has never struck me as a particularly durable competitive game, at least not when owning that infrastructure isn’t strictly necessary in the first place. Owning the warehouses is the expensive path, because then you are the one holding inventory that sits idle and goes obsolete. Historically, when industries mature, companies tend to stop trying to own every link in the chain and instead specialize in their core competency. Even Coca-Cola doesn’t want to bottle its own soda anymore, it wants to be capital-light and sell the syrup, letting bottlers handle the capital-intensive part.

Apple doesn’t try to be a logistics company. FedEx and UPS exist for that. If you pour endless capital into building out more warehouses and delivery infrastructure without proportionally accelerating the volume of goods flowing through it, you end up as a capital-intensive business generating mediocre returns on assets, equity, and invested capital in general.

This is also why JD cannot simply copy PDD from where it stands. JD is a warehousing and logistics operation that happens to run an ecommerce platform on top of it. Its model runs the other direction, and demand aggregation is not what it was built around, so it cannot bolt PDD’s model on without unwinding the very assets that define it. These are different businesses with different economics, and one of them does not become the other just by deciding to.

That capital-light positioning is also where PDD’s biggest structural vulnerability comes from.

Built on WeChat’s Back

This capital-light advantage comes with real structural vulnerability, and it traces back to PDD’s origin story. PDD is essentially a tenant living inside Tencent’s WeChat ecosystem, a brilliant arrangement but increasingly a liability, especially now that Tencent, a major PDD shareholder, is once again trying to build a competing shopping experience directly inside WeChat itself.

The origin story explains why. Alibaba’s rivalry with Tencent meant WeChat blocked Alibaba links entirely, and Alibaba already had Alipay and wanted to stay independent anyway. JD’s case is different. Tencent invested in JD back in 2014 and handed it the most prominent shopping placement WeChat has ever given anyone, a Level 1 access point wired straight into the app, the only e-commerce company ever to get that kind of positioning. But JD ran that placement as a conventional storefront, a direct-sales link that behaved like a smaller version of its own site. It never turned WeChat’s social network into a growth engine. That was the real gap, and it’s the one PDD walked into. Nobody had taken the sharing mechanic itself, the act of forwarding a deal to friends and family, and built an entire shopping model on top of it. Since nearly everyone in China uses WeChat, PDD didn’t even need its own app initially. It launched as a simple WeChat message: “Want a box of fruit? Share this with friends, and once enough people join, we’ll ship it.” That was the entire distribution mechanism, an ad that spread virally through personal networks. When WeChat later introduced mini-programs, lightweight apps that run inside WeChat itself, PDD built its full platform on top of that infrastructure.

So while Tencent is now trying to compete with its own tenant, it’s not as simple as flipping a switch. Users have built genuine habits, opening PDD, watering their cartoon tree, running through the gamified routines, that don’t transfer easily to a new interface, even one hosted by the same underlying app.

If the consumer-side moat is habit, the supplier-side moat is something even harder to replicate.

The Supplier-Side Network Effect

The less obvious moat is on the supplier side. PDD spent roughly a decade building direct relationships with manufacturers, and one of their sharpest insights was recognizing how much idle factory capacity exists between seasonal demand cycles. PDD’s pitch to these manufacturers was essentially: post a group-buy listing, and once enough orders stack up, produce and ship immediately, no inventory risk, no upfront cost, orders already guaranteed before a single unit is made.

That solved a real problem for factories sitting idle outside peak season, and in doing so, PDD made itself structurally necessary to its supplier base, not just as a sales channel but as the data layer that understands consumer demand patterns in real time. That’s a network effect comparably as durable as owning a logistics network, except it costs almost nothing in capital to maintain. Combine that with the consumer-side habit loop, and PDD has built a two-sided flywheel: manufacturers depend on PDD’s demand-matching, and consumers depend on PDD’s habitual, gamified shopping experience. Both sides reinforce each other, and neither requires PDD to own a single warehouse.

All of that is the qualitative case. Now for the numbers.

When the Numbers Do the Talking

All the reasoning above explains why PDD’s model works, and that reasoning is important, it shouldn’t be waved away as secondary. But at some point, the qualitative case has to show up in the numbers, and this is where PDD’s story becomes genuinely hard to believe.

What should jump out from that table more than anything else is the sheer explosiveness of PDD’s growth trajectory relative to its two much older rivals. When PDD launched, Alibaba was already the dominant, entrenched e-commerce platform in China, backed by a decade-plus head start, a mature logistics ecosystem, and Alipay’s grip on Chinese digital payments. For a company with no app, no warehouses, and no owned logistics network to even approach that kind of scale within a decade is already improbable.

In November 2023, PDD briefly surpassed Alibaba in market capitalization for the first time, a moment that stunned much of the investing world given the two companies’ starting points a decade apart. That lead has not held consistently. Alibaba’s market cap has since moved back above PDD’s, with Alibaba around $230 billion and PDD around $120 billion as of mid-2026. But the fact that a scrappy, capital-light, gamified marketplace was ever seriously in the conversation with Alibaba at all is the real point.

Put this in Western terms, and the analogy becomes almost absurd. Imagine a company nobody had heard of a decade ago briefly overtaking Amazon in market capitalization for e-commerce dominance. Nobody in the Western investing world would have found that plausible on paper, yet that’s roughly the scale of what happened in China’s market between PDD and Alibaba.

Why It Should Catch Your Eye

What should catch your eye about PDD in the first place is really simple: it’s trading at an absurdly cheap valuation relative to the cash generation underneath it.

As of early July 2026, PDD carries a market capitalization of roughly $117 to $120 billion, well off its highs after a rough stretch of earnings misses and analyst downgrades through the first half of 2026. Against that market cap, the FY2025 results show RMB422.3 billion, about $60.4 billion, in cash, cash equivalents, and short-term investments sitting on the balance sheet. You’re buying the underlying business for an enterprise value of roughly $57 to $60 billion.

Now look at what that enterprise value is actually buying. FY2025 revenue came in at RMB431.8 billion ($61.8 billion), up 9.7% year over year, and net income was RMB99.4 billion (about $14.2 billion). More importantly for a cash-flow lens, PDD generated RMB106.9 billion ($15.3 billion) in operating cash flow for the year. Even using operating cash flow as a conservative stand-in for free cash flow, given how asset-light PDD’s model is with minimal capital expenditure needs, that puts the enterprise value at roughly four times operating cash flow. And that multiple is calculated before counting another RMB100 billion or so in long-term time deposits and securities, plus the restricted cash tied to the payments business. Count those and the enterprise value shrinks further still.

So why the heck is it trading so cheap?