The Battle of Uncovering Fake Accounts
The war of the good with the bad goes a long way back in history. In the technology world of today, this is no different with bad actors taking center stage, albeit a hidden way, to cause financial losses on a large scale and erode digital trust. Fake accounts are a huge challenge faced by all social and marketplace companies today.
The first step for fraudsters to commit an intended fraud on a given platform begins with fake account creation. The more accounts they can create, the more is their ability to cause damage. To keep the customer experience smooth, most online platforms make account creation a simple process. This makes the business platform an easy prey to get compromised at the hands of bad actors. Moreover, fraudsters, spammers, and scammers who create fake accounts are becoming sophisticated in both technique and magnitude, making it difficult and time consuming to detect accounts that are not legitimate, and often, after the damage is done.
Why Should Businesses Care?
A business might believe that accounts appearing to be harmless will not impact it in any way. However, such a belief can be damaging as many innocuous accounts may sit harmless for months and become a fertile ground for subsequent criminal behaviour. It is important that these be removed or blocked. By removing or blocking fake accounts, businesses can empower themselves and benefit from providing online safety of their valuable customers, preserving brand reputation, and measuring accurate user growth.
Online safety of valuable customers
A fake account is a means to an end. By detecting fake profiles, businesses can prevent downstream damage. In most social and marketplace platforms, fake accounts result in fake likes, votes, friending or following. This also enables a fraudster to commit promotional abuse, phishing, spamming, rendering fake reviews or acting as an imposter to fraud legitimate users. Exposing a genuine user to different types of this fraud impacts their safety if they want to perform any action within the platform and reduces their trust and reliability.
Preserve brand reputation
Using fake accounts, the fraudster can greatly damage the reputation of the platform by impacting customer experience and customer engagement. Negative or inappropriate content can affect the brand of a growing company, which could deter new customers from using the platform, which in turn could affect customer growth, brand and revenue.
Measure user growth
Growth is a function of product marketing and network effect. Many companies use analytics to measure the performance of their marketing campaign or how they want to spend their resources — monetarily and otherwise — based on the growth they see in specific regions. With fake accounts, this picture will not be accurate. To make sure they are investing in right areas as part of their company’s strategy, it is very important to have the real picture of the growth, which is possible by blocking or removing fake accounts.
Fake Account Creation
It is beneficial for fraudsters to create multiple accounts in one go. This gives them an ability to use other fake accounts if one is caught and blocked. It also enhances scale of the attack. Moreover, the effort to create one or more is the same. It is also very cheap to create fake accounts using bots/device farms. Bots account for 21.8% of the overall internet account.
Given that most fraudsters want to create more than one fake account at a time, the methods they use are quite creative and sophisticated. Due to the high sophistication of attack types and obfuscation techniques, it is very hard to detect fake accounts consistently.
DataVisor Solution for Account Abuse Prevention
It is hard to detect fake accounts with the regular rules-based engine and manual reviews are not helpful. While there are machine learning solutions in the market claiming to address this problem, the ever changing behavior of the fraudsters makes it hard to detect them. DataVisor’s singularly unique approach addresses these challenges head-on delivering faster results with better coverage and accuracy with reduced manual review. Here are reasons why:
Results on Day 1, No Retuning Needed
General supervised machine learning or rules require a long time to learn from the historical labels or patterns to catch similar fraud. They cannot adapt to the constantly evolving attack techniques. Fraud patterns for account creation are frequently changing and this necessitates frequent retuning. DataVisor’s patented algorithm does not require labels or retuning, and adapts to the changing patterns without relying on previous signals to detect the new attack pattern. This means results are available on first day itself with little to no maintenance.
Better Coverage and Accuracy
DataVisor’s algorithm is able look at a fraudster not just individually but also in a group. Since most of the fake accounts get created in bulk, this solution, built to find groups of fake accounts with correlation, can identify such accounts easily. By finding the entire network together, there is a higher likelihood of detecting suspicious behavior — and taking timely action. When users share similar devices,ips etc. by premise itself, the technology identifies that they are correlated, hence a better justification to the accuracy of the results. DataVisor delivers greater coverage and accuracy than most of the solutions in the market today.
Reduced Manual Review
Because of the high coverage and accuracy of the DataVisor solution, based on the score, a customer is able to automatically block such accounts thus helping to reduce the risk as well as manual review. Lower manual review means reduction in costs and better automation.
Organizations that want to defend themselves from account abuse have superior and reliable technologies available at hand today to grow smoothly and with confidence.