Many companies in the consumer goods sector, especially in electronics, encounter many problems with the online sale of counterfeit products. Managing reputation is difficult enough but the problem of fake or counterfeit goods has been growing rapidly in the last few years with the growth of e-commerce platforms and the explosion in online sales. With complex supply chains and multichannel sales keeping track of every single route to market is incredibly difficult.
Counterfeit and fake goods are available not only through unauthorized online channels but even the large legacy sites like Amazon are used for counterfeit products. In this case, we describe how Natural Language Processing (NLP) techniques can be used to trace and attack counterfeits on e-commerce platforms.
Why should you read this case study?
Tracking counterfeit and fake goods is a complex difficult task. It can be a major challenge for those working brand management, sales channels and marketing as well as legal teams.
Is counterfeiting something you fight every day? Are you side by side with your legal department talking with your resellers to hunt after counterfeits? Are you working in brand management in a sector where counterfeiting may affect your brand reputation? Do you work in a complex multi channel environment where customers are often faced with counterfeit goods? Read this article and we will see how customer feedback analysis can allow you to be more effective in your work in controlling and managing your brand.
How to find counterfeits through customer feedback?
Using NLP (Natural Language Processing) techniques to trace counterfeits is a completely new approach. Nevertheless, Wonderflow has taken on the challenge to do so for one of its clients that produce a large range of electronic devices for consumers.
Wonderflow’s first challenge was to run extensive analysis on how to trace the specific reviews that refer to counterfeits. Our NLP engine scanned several e-commerce platforms with the aim of finding reviews where buyers seem to complain about fake products. After uploading and cleaning the data, The Wonderflow AI trained itself to understand the kind of language and sentiment customers were using which would refer to a counterfeit product.
Through identifying the purchase channel we were able to identify that in some cases up to 70% of the products sold on a specific website, were fake! So with the help of buyers and their feedback, Wonderflow was able to find many instances of fake products being sold as legitimate, cheating not just our client but hundreds or even thousands of customers as well.
The challenges we faced:
- Extensive analysis to find the right topics and language used by customers when referring to fake products they bought
- Establish cooperation from the independent e-commerce platforms to delete traced counterfeit goods*
- Make sure all activity was done within the legal boundaries
* After identifying this large group of fake listings, the e-commerce platform was not willing to remove the products from their website. Only when Wonderflow’s client company filed a lawsuit did they begin to cooperate with the company. Until then, not only was the brand being damaged, but buyers were losing their money.
You may imitate, but never counterfeit
- The company is now able to decrease counterfeits being online within 90 days through proactive analysis
- Enormous financial upside for the client. A conservative calculation showed that the company saved itself 15 million USD in missed sales as well as the cost associated with less brand damage per year
- Huge advantage for e-commerce reseller in fewer returns, less commercial loss, building consumer trust through supplying real brands
The information in this case study is confidential. Brands and company names are anonymized.