‘Brushing’ casts doubt on Alibaba figures as $20bn IPO looms

After four years managing a private delivery company in the Chinese city of Ningbo, Chen Qian has acquired a new skill: he can tell which packets are fake even before he picks them up. Some are hollow boxes, some rattle with a piece of candy or a keychain. Recently, he says, merchants sending fake deliveries have started putting toilet paper rolls to give some heft.

Mr Chen says these account for about a quarter of the 4,000 packages his company handles every day. The phenomenon is widespread throughout China; a consequence of the country’s booming e-commerce industry and, specifically, a practice known as shuaxiaoliang, or literally – “sales brushing”. Online sellers are recruiting their friends, relatives and even professional fraudsters to make fake orders because shipping more goods would give them better placement – and therefore a better chance to garner more real sales – on websites such as Alibaba-owned Taobao.

In some category of goods, fake sales can account for between a 10th to a quarter of all online sales, according to a series of interviews with ecommerce vendors, logistics companies, and people who help fake internet traffic for e-commerce sellers.

This high proportion calls into question the key operational metrics published by Alibaba ahead of its expected New York listing this month, when it is likely to raise around $20bn and eclipse Facebook and Google to become the biggest ever internet IPO.

Since Alibaba’s Tmall and Taobao sites account for 80 per cent of the overall online retail volume in China, “brushing” also calls into doubt China’s official e-commerce statistics.

Alibaba said in a filing to the Securities and Exchanges Commission this week that it handled $296bn worth of goods, consisting of 14.5bn orders, in the year ended June 30. Ebay handled $81bn worth of goods over the same period.

Alibaba noted in the risk factors section of its prospectus that sellers on its site may “engage in fictitious or phantom transactions with themselves or collaborators in order to artificially inflate their own ratings on our marketplaces, reputation and search results rankings”.

The company declined to comment further due to a pre-IPO silent period, but Alibaba has been cracking down on brushing for the last three years. This has had some effect, according to sellers and others active on Taoboao, although the practice still flourishes as sellers stay ahead of Alibaba’s audit methodology.

Brushing highlights the headaches of policing third parties on e-commerce sites and applies not just to Alibaba but to all sites that have open supplier platforms, such as Ebay and Amazon. However, Alibaba is most affected due to its sheer size in the Chinese market, and because fierce competition and rising advertising rates charged by Alibaba mean that the vast majority of sellers on Taobao are now lossmaking. Ebay and Amazon declined to comment.

Zhang Yi, chief executive of iMedia Research, a mobile internet consulting group, said a private study by his group reckons that a very large number of shop owners on Alibaba’s flagship ecommerce site Taobao, which accounts for two-thirds of Alibaba’s total sales volume, “brushed” in the first half of 2014. “The great majority are brushing or have ‘brushed’ at some point” he said.

“We’ve only started brushing recently,” said one Taobao shop owner in Hangzhou which sells hats and traditional silk scarves, who asked not to be identified. “There is no other choice for us. A lot of the other shops have been doing this for years, and we realised that no matter how well we did in sales, we could not compete with those who brushed. The way I see it, it would be best for all of us if nobody brushed.”

“But the competition is so fierce, there is really no other way, all our competitors are doing it. If your sales aren’t high enough, you will get low placement and no sales”

There is some disagreement over the extent to which brushing affects Alibaba’s overall sales numbers. One person familiar with the company said Alibaba was aware of the problem but did not consider it material enough to impact overall sales.

However Anne Stevenson-Yang, head of J Capital Research, the Beijing based economic research group, drew attention to the 63 per cent jump in Alibaba’s gross merchandise value between the end of 2012 and end 2013, from $157bn to $248bn.

In that same period, she says, there was only a 6 per cent revenue growth for all retailers listed on the Shanghai, Shenzhen, and Hong Kong stock exchanges, not corrected for mergers and non-core investments.

“You can’t say this is channel migration, because a lot of the consumer/retail companies have very robust online sales, and those sales are growing more slowly than offline sales” she says. “Manufacturers of products sold online are seeing slow or negative growth. So where are all these online sales coming from?”

•••

Over the past three years, Alibaba has improved its auditing procedures, which use algorithms to determine suspicious activity. Merchants caught brushing could be downgraded or even kicked off the website.

Mr Chen, of the delivery company, told the FT via telephone that the number of empty packages he handles has been reduced from half to one-quarter of the total in that time. However, the practice still flourishes as sellers stay ahead of Alibaba’s audit methodology. In chat rooms and blog forums, vendors discuss how not to get caught: do not have too high a conversion ratio of sales to internet traffic clicks, and do not “brush” from the same IP address too often.

Li Siyuan, who sells cut flowers online on Taobao from his Beijing flat, said he personally does not brush sales, however he knows many merchants who do, but the practice is declining. “The golden age of brushing was 2009 to 2011” he said. Nowadays he estimates that the total amount of fake sales in the online flower industry is 10 per cent or less. “Today the best way to get high sales is to offer a great product,” he adds.

But Mr Chen, of the delivery company, says it will be difficult to completely eradicate the practice. “Even though our employees can pretty much tell which ones are empty, the line is still rather blurry,” he says. “Our clients can insist that they just intend to send a pack tissue paper; or even harder still, if they send out a receipt in an envelope.”

It appears that brushing violates no laws and, arguably, benefits everyone – store owners get better listings, logistics and delivery companies get more sales, and Alibaba gets a boost in traffic.

“I don’t think I want to criticise the practice too much because we get so many sales,” says Mr Chen.

Additional reporting by Ma Fangjing in Beijing

Brushing: how it works

‘Brushing’ essentially consists of creating fake orders – the merchant sends out an empty box or delivery envelope accordingly, but refunds the money paid by the ‘purchaser’.

The practice of shipping an empty box or entering a fake order code is necessary because Alibaba requires a unique delivery code to be entered with each order.

In practice, Chen Xujie, a man from Wenzhou who fakes internet traffic for ecommerce sellers on his website 668shua.com, says half of the fakery is done by shipping empty parcels, and half is done via a grey market in active order codes sold by logistics companies to vendors via specialised websites.

Brushing has generated a whole series of side industries in China. Mr Chen, for example, runs a thriving business in artificially boosting online traffic to Taobao and Tmall shops. Shop owners who brush but do not also fake their traffic numbers can get caught because they would appear to be too successful, which casts suspicion and can cause them to be automatically downgraded by Taobao.

Some sellers also use virtual private networks on computers to fake different IP addresses for order locations, he said.

Going by the online chat handle of “Stupid Jerk”, Mr Chen uses internet bots and software to fake traffic. “Nine out of every 10 sites on Taobao do it” he said via instant messenger.

He claims to make Rmb3,000 per month generating fake traffic for Taobao sellers, and says he got into the practice after owning a shop on Taobao selling Korean cosmetics. He left Taobao because he said “there was no hope”.

“I spent too much time brushing my sales and it still wasn’t enough”.

Alibaba Valuation Calculator

This interactive calculator uses a basic two-step discounted cash flow model to help illustrate how variations in key assumptions can change the potential market value and share price of Alibaba. It is not an accurate valuation tool.

The calculation assumes that Alibaba will be paying around 12 per cent tax every year on its earnings.

By Tom Pearson, Robert Armstrong, Sujeet Indap, Steve Bernard and Robin Kwong