Flanked by a wall of vivid orange packages, YouTuber Hope Allen was uncertain about the phenomenon they represented. “Temu has been taking over, and I don’t know how I feel about it.”

The ecommerce firm’s “shop like a billionaire” ad campaign to promote its online flea market was ubiquitous. Allen’s response was to buy a job-lot of undeliverable parcels that disgorged a sample of Americans’ experiment with Temu.

Among the piles of extremely cheap clothes, bags, tools, toys and kitchen implements were designer look-alikes that provoke questions. “How do they get away with that?” asked Allen of a faux Victoria’s Secret bag, before recoiling from the smell.

All were flown in from China by Temu’s parent company PDD Holdings, a self-described “agricultural group” engaged in what might be the fastest and most ambitious retail expansion in history.

With Temu, PDD wants nothing less than to change the way the world shops, a faster, leaner and cheaper version of Amazon that has spread from China to 49 countries after less than two years in operation.

The plan, as best can be inferred, is to use blanket advertising to lure western consumers to Temu’s app, where algorithms and AI anticipates their whims and desires. Products are shipped for free direct from China’s factory gates, cutting out the middleman and ensuring low prices.

Temu’s sister app Pinduoduo already dominates China. When it still published such numbers, PDD reported more than 870mn active users in the country supplied by over 13mn merchants who, it claimed, together generated a third of all parcel traffic in the country, tens of billions of packages a year.

After just nine years in business, PDD is now bearing down on the world’s biggest ecommerce group Alibaba, both in terms of retail scale and stock market capitalisation. Worth $162bn, it regularly trades places with the older retail giant as the most valuable Chinese company listed on a US stock exchange.

But the tale told by those incredible numbers is a mystery that raises more questions than it answers.

For instance, why does PDD look like much smaller peers when staff levels and research spending are compared? Why haven’t competitors described the impact of PDD’s rise? Why do balance sheet metrics move at a different pace to revenues? How does a $200bn company own less than $150mn worth of hard assets?

These questions add up to a bigger one: why do US investors have so much confidence in an opaque operation whose financial statements seem to lack much pattern or explanation, and whose operations, management, auditors and regulators sit in a distant untouchable jurisdiction?

A PDD spokesperson said: “We respectfully disagree with that characterisation.” They encouraged the Financial Times “to review our financial reports and earnings calls for comprehensive insights” and said “disclosure rules and fairness principles” prevented it commenting on those financials to “individual media outlets”.

The American stock market is scattered with cautionary tales of Chinese promise that ends up unmet. “I drank a lot of coffee in a lot of Luckin Coffee locations, and it was still a fraud,” says one investor, of the once $13bn Nasdaq-stock which admitted to inflating sales figures in 2020. Wish, an algorithmic Temu-like ecommerce site, saw a $22bn market value dwindle almost to nothing.

Yet few have promised as much to investors as PDD. It operates like eBay and Amazon’s third party marketplace, connecting buyers with sellers to take a cut of each transaction and charging merchants to advertise on its platform.

In its most recent quarter those revenues almost doubled versus the previous year, to $9.4bn, prompting Alibaba founder Jack Ma to exhort his former company to “change and reform” in response. PDD reported $2.5bn of cash flow, even as it appears to throw very large sums at the expansion of Temu.

It has achieved this with a headcount that upends all assumptions about ecommerce logistics: it started last year with 12,992 employees, an order of magnitude less than Alibaba and a small fraction of Amazon’s 1.5mn staff.

PDD’s physical footprint is also minuscule, a striking contrast with Amazon, JD.com and Alibaba, where control of logistics was long seen as a competitive advantage; a way to ensure speed, capacity and satisfactory service.

Where Alibaba spends $5bn a year on property and equipment, including the upkeep of 1,100 warehouses, PDD owns just $146mn of hard assets — mainly office equipment and IT hardware and software.

It didn’t disclose any leases of warehouses before 2021, when it said its online grocery business, then just one year old, had expanded to serve more than 300 major cities in China. It doesn’t report the size, location or number of the warehouses it rents.

Those logistics, like PDD’s servers and customer service call centres, are mostly outsourced, ephemeral and unenumerated. The opacity extends inside the business. Staff use pseudonyms and know little about other teams. The structure is flat, with a small group of decision makers directing the “grassroots”, young people chosen for their poverty or debt obligations which motivate them to work long hours. PDD said: “Employees have access to all necessary information for effective cross-team collaboration and to fulfil their roles.”

Such secrecy is a hallmark of Colin Huang, a former Google engineer who founded Pinduoduo in 2015 and for a long time hid his ownership. Initially, it attracted users by gamifying commerce. Describing itself for a while as “Costco + Disney”, Pinduoduo offered games that mimicked addictive titles like Farmville and Candy Crush.

Combined with micropayments, carefully calibrated rewards, coupons and offers, the software was said to keep consumers returning to the app, where they were tempted to make impulse purchases or promote Pinduoduo to their friends. By 2018, the app boasted 300mn customers and after only three years in business the company listed on the Nasdaq stock exchange, raising $1.7bn.

PDD vice-president of strategy David Liu, a former Goldman Sachs banker, described its “push-based business model” on a 2019 podcast. He said that instead of searching for particular items, people went to Pinduoduo to browse, where algorithms and highly targeted offers encouraged social sharing of purchases. The rise of mobile internet use, according to co-chief executive officer Chen Lei, created a “new paradigm.”

Their descriptions are hard to square with the numbers. Almost all of PDD’s revenues then came from marketing services, where merchants “bid for keywords that match product listings appearing in search or browser results”.

Another claimed benefit were insights into consumer demand that PDD passed on to merchants. As an example of this cutting-edge information, Liu talked about a glassware manufacturer that supplied big western retailers: PDD helped it to understand that Chinese people liked stubby wine glasses to go in their small cupboards.

The experiment’s end may have been overshadowed by events. Founder Huang, one of China’s richest people, stepped down in 2021 declaring a desire to research food science and life science. His letter said: “I would feel very lucky and blessed if I have the chance to become a research assistant to a future, possibly great, scientist.”

He was replaced by co-chief executives, Chen and Jiazhen Zhao, who as a sideline personally control a payment services provider to the company that they purchased with the help of a $100m loan from PDD.

A PDD spokesperson said this was “a simpler legal arrangement to secure and complete the acquisition of a payment licence.”

Straight after Huang left PDD made a long-awaited flip to profitability. Having lost Rmb27bn ($4.2bn) in three years, PDD reported earnings of Rmb2.4bn ($380mn) for the second quarter of 2021. It was a key moment, as the group had raised $11.6bn from sales of stock and debt that converts into shares to support a spend-now-for-profits-later business model.

Rather than focus on this milestone in the August 2021 earnings call, Chen described PDD’s efforts to support Henan province after it was hit by heavy rain.

More unusual still, Chen then announced a plan for the first Rmb10bn of profits to be generated by PDD. He intended to spend it all on an “agricultural initiative” without a clear business purpose. The aim was fuzzy and broad, ”to facilitate advancement of agritech, promote digital inclusion, and provide agritech talents and workers with greater motivation and a sense of achievement”, he said.