阿里起訴淘寶假貨店主，攜手數家公司利用大數據打假

電子商務能夠稱之爲過去十年間人們生活中最大的改變。如今咱們能夠在下午5點下單，次日早上9點就能收到貨，這在以前是很難想象的，由於以前人們一般都要花好幾周的時間來找本身想買的東西。

電子商務的興起也帶動了新經濟的發展，成千上萬的商家都經過電子商務平臺直接將商品出售給消費者。甚至還有不少人辭職後專門作電商，月薪高達六位數。

可是，電商的發展也致使了一些很差現象的發生，特別是假冒僞劣商品的肆虐。英國有個叫阿密特·沙馬的人，經過在eBay賣假名牌服裝，四年獲利超過100萬英鎊，由此被判處21個月的監禁。

電商賣假貨的問題並不僅在美國和歐洲國家出現，在中國，阿里巴巴日前也起訴了兩名假貨販賣商，他們經過淘寶銷售假冒的施華洛世奇手錶，阿里巴巴要求他們賠償人民幣140萬元。

質量問題也是電商行業的一大弊病，以次充好的狀況很是嚴重。不少商家會買水軍，增長商品的好評，並提升商家等級。這樣一來，消費者購物時就會覺得這個店裏的商品或服務已經獲得了不少消費者的承認。而事實上並不是如此，商家只是花了點錢找了水軍而已。

這給不少電子商務網站形成了不少的困擾，由於平臺自己沒法干預消費者對商家的評分。

若是你想買一塊手錶，在實體店裏你能夠先試戴，檢查各部分是否無缺無損，看看是不是正品，可是在電商平臺就不能夠了。若是人們在電商購物時遭遇了以次充好的狀況，那麼他們極可能就會迴歸傳統的實體店購物方式。因此這個問題不獲得解決，電子商務將流失大量顧客。

若是你花了50塊錢買了坑爹貨，正常人都不會願意再花500塊錢在電商平臺買東西了。因此如今電商平臺上不少價格較高的商品都很差賣，可是商家還要囤着貨，這樣一來不少商家都浪費了大量的儲貨空間，甚至還要在這些商品的安保上花費更多。

不少比較實誠的賣家都退出了電商平臺，由於不少店家經過買水軍的方式提升店鋪等級，以這種方式惡性競爭。

在這些問題更加猖獗以前，咱們必需要採起行動了。阿里巴巴已經攜手數家公司聯手打擊假冒僞劣產品。阿里巴巴集團與路易威登、三星和瑪氏公司達成合做協議，利用大數據找出淘寶平臺上的假冒僞劣商品。

阿里巴巴集團首席平臺官傑西·鄭認爲，大數據和分析是當今打擊假貨的最有力武器，各方必須聯合起來共同行動，她說「咱們的大數據分析能力很是出色，因此咱們有信心這次合做可以有效的改變假貨猖獗的狀況。」

阿里巴巴與這些公司合做，阻止假貨繼續傷害它們的品牌。阿里巴巴自己打擊假貨的能力也是很是值得確定的，在2016年4月至7月間，阿里巴巴沒收了價值2.072億的假冒僞劣商品，關閉了417家非法交易假貨的店鋪，幫助警方抓獲了332名造假犯罪嫌疑人。

經過機器學習，阿里巴巴的系統天天都會檢查1000多萬產品清單，2015年8月至2016年8月間，累計剔除了38000萬產品，取消了180,000個第三方賣家的銷售資格。

這次阿里巴巴與各公司的合做將更加全面、有效地打擊假貨。阿里巴巴對假貨問題的關注度之高可能會令外界震驚，在他們2016年年度報告中，「假貨」一詞出現了30次之多。

阿里巴巴並非惟一遭遇假貨問題的電商平臺，亞馬遜上的假貨問題一樣很是突出。

舉例來講，此前蘋果公司起訴了亞馬遜上的一個店鋪，該店鋪出售的蘋果產品90%都是假貨，一旦這些假手機運行出現問題，亞馬遜和蘋果公司的聲譽都會受到傷害。一些造假者甚至買通了亞馬遜後勤部，讓他們的假貨也能進入亞馬遜倉庫，讓消費者覺得他們買到的是正品。

亞馬遜對此很是頭疼，他們在簡化交易的同時，也讓不少假冒僞劣商品趁機佔領了購物網站。爲了阻止這些假貨販賣商，亞馬遜一樣採用了大數據，提升第三方賣家銷售大品牌產品的門檻。關於這一舉措，亞馬遜並無公佈太多細節，他們宣稱已經在這個行動中投入了數千萬美圓。他們同時表示這一過程還須要各大品牌的參與合做。

阿里巴巴和亞馬遜開始整治假貨並不在人們意料以外，再不行動恐怕就要讓他們陷入大麻煩了，畢竟法院和媒體都已經關注它們好久了。

蘋果公司起訴了亞馬遜上的Mobile Star店鋪，該店鋪經過亞馬遜銷售的蘋果產品90%都是假貨。儘管此案的被告並非亞馬遜，可是這對亞馬遜來講也不是什麼光彩的事情。還有一些其餘的小品牌也曾公開指責亞馬遜的假貨問題，如TRX（美國健身器材公司）和勃肯（德國鞋業品牌），勃肯已經宣佈從2017年1月1日起全面撤出亞馬遜。

亞馬遜確實是挺倒黴的，龐大的規模是他們的優點，但如今卻讓他們陷入困境。

2017年1月份亞馬遜庫存商品超過39800萬件，較2016年12月增長了8%。在如此龐大的商品數量面前，光靠人工是不可能篩選出假貨的。不少賣家都抱怨稱亞馬遜以前的系統只能在接到假貨舉報後才能處理銷售假貨的商家，可是這些受處處理的不法分子下次換個店鋪名還能接着賣假貨。因此僅僅靠舉報系統和客服團隊是不可能處理這麼多假貨的。

隨着技術的日益成熟，大數據、機器學習和人工智能進一步發展，這些電商平臺將能夠更快速且高效地打擊假貨。亞馬遜如今迫切須要認識到這一點，也許還要從他們的中國競爭對手——阿里巴巴身上多多學習。

 

英文原文

 

Can Big Data Stop The Shadiest Elements Of E-Commerce?

All is not well with many e-commerce companies, but could big data help?

E-commerce has been perhaps the single biggest change in people’s habits over the past decade. Now we can realize that we need something totally obscure at 5pm and have it in our hands by 9am the next morning, where previously we would have needed to search for weeks just to find it. It has also created a new economy, where there are now thousands of people who can sell their own products directly to consumers through e-commerce platforms. There have even been stories of people making six figure salaries by selling through these kinds of platforms instead of holding down regular jobs.

However, the opportunities that it has given honest people has also led to a considerable rise in the number of counterfeit and low quality products available. There have been stories like that of Amit Sharma in the UK, who earned over £1 million in 4 years selling counterfeit clothing on eBay and was jailed for 21 months after being caught. It is not just a problem for US and European countries either, with Alibaba suing Liu Huajun and Wang Shenyi for 1.4 million yuan for ‘violation of contract and goodwill’ after they were found to have been using Taobao to sell fake Swarovski watches.

Similarly there have been issues with low quality goods being sold as considerably higher quality than they are. This is achievable because many online sellers have realized that they can utilize companies who offer to increase the user ratings for products and seller accounts on e-commerce sites. This means that people who buy these products normally do so under the impression that hundreds of customers have been happy with the quality of the product and service offered by the seller. In reality they may have simply paid a company to increase their user ratings.

This represents a serious issue for e-commerce sites, as it abuses the one thing that they can never offer - a preview.

In a brick and mortar store, if you want to buy a watch, you can try it on, check that everything is working properly and check the authenticity. Nobody will ever be able to do this with the majority of e-commerce companies, so if people are a victim of either of these problems they are far more likely to be driven back to traditional forms of shopping experience. It also has an impact on the amount people are likely to spend online. After all if you get burnt buying a $50 watch, you are never going to risk it with a $500 one. This creates a situation where e-commerce sites will struggle when it comes to selling big ticket items, but will still need to stock them - so excess stock will be held by the company, wasting warehouse space and likely requiring additional security.

This then causes issues for genuine sellers who refuse to use these kinds of services, because traditionally the companies with the most highly rated reviews are the ones who appear at the top of searches.

However, there are moves being made to prevent these issues from becoming unmanageable, with Alibaba teaming up with several companies to try and combat counterfeiters. The company is going to be working with a selection of companies including Louis Vuitton, Samsung and Mars to utilize big data and identify fake products on their platforms. Alibaba have great faith in the move with their chief platform officer, Jessie Zheng claiming ‘The most powerful weapon against counterfeiting today is data and analytics, and the only way we can win this war is to unite…With our robust data capabilities, we are confident the alliance will accelerate the digital transformation in our global fight against counterfeits.’

It is likely that the collaboration with companies with a vested interest in stopping these counterfeiters and damaging their brands. It is likely to bring even more robustness to Alibaba’s already reasonably robust approach, having already seized $207.2 million of counterfeit goods, shut down 417 production rackets and helped to arrest 332 counterfeiting suspects between April and July 2016.

Using machine learning, Alibaba’s system scans 10 million product listings every day and had removed 380 million product listings and 180,000 third party sellers in the 12 months leading up to August 2016. It is hoped that by bringing together every stakeholder impacted by fake products this process could become even more thorough and effective. It is little surprise that this has been such a focus for the company given that some form of the word ‘counterfeit’ appears 30 times in Alibaba’s 2016 annual report.

Alibaba aren’t the only company suffering from this phenomenon though, with Amazon being a highly visible target for counterfeiters. For instance, according to a lawsuit filed by Apple, 90% of Apple products (mainly chargers and peripherals) sold on Amazon are counterfeit, meaning that Amazon’s reputation and Apple’s reputation is damaged when these fake goods malfunction or break. The fraudsters are even taking advantage of the logistics offered by Amazon, allowing them to send their fake products directly to an Amazon warehouse, giving them an air of authenticity. It has created a major headache for Amazon, who’s attempts to make selling through their marketplace as simple as possible has led to a huge number of fraudulent products flooding the site. To prevent this, they are also looking to utilize big data with their partners, as well as making it more difficult for third party sellers to sell big brand products. The public details of this are limited, although Amazon have said they spend ‘tens of millions of dollars’ on the endeavour. They also say that they need to work with brands in order to make the process work.

It is little surprise that both Alibaba and Amazon are suddenly taking this move seriously, because it is beginning to hit them where it hurts - in the courtroom and in the media.

For instance, Apple filed a lawsuit against Mobile Star who sold Apple products through Amazon after it found that 90% of the Apple merchandise sold by the company was counterfeit. Although this was not a lawsuit against Amazon, having the world’s most valuable company suing somebody for something they did through your marketplace is not good for business. There are also multiple examples of smaller companies like TRX and Birkenstock openly criticizing Amazon for their approach, with Birkenstock actively withdrawing all products from Amazon as of January 1 2017.

In many ways you have to feel sorry for Amazon, as it seems that their biggest strength has also turned into their biggest weakness - their size.

Amazon stocks over 398 million products as of January 2017, with an 8% increase from December 2016 alone. Taking a ‘human first’ approach would be impossible in the face of these kinds of numbers. Many of the sellers who have complained about counterfeiters have needed to rely on a reporting system where they can report to Amazon who then take down the counterfeiter’s page, only for the same seller to appear under a different name days later. This could happen to any one of the nearly 400m products, so trying to manage this with a report system and complaints team would be impossible.

As technologies progress and the use of data, machine learning and AI improves even further, the chances of stopping this more quickly and effectively is going to increase too. It is something that Amazon is desperate need of and perhaps they should be taking the cues from their biggest Chinese rival.

 

轉自：燈塔大數據；微信：DTbigdata