Вы используете устаревший браузер Установите более современный ¯\_(ツ)_/¯
Driving business

Retail Fraud: Losses from Rewards


Artificial intelligence is a great tool to discover and address fraudulent schemes that target loyalty programs in online retail.

Why customer loyalty programs fail to pay off?

Say, an online retailer invests in the development of a customer loyalty program, implements it, and then suddenly finds out that brings a negligible return, if any. Such cases are not uncommon.

Rewards accrue, and purchases are sometimes made, but the number of new loyal customers is disappointing.  Was there a miscalculation by marketing managers? Not necessarily. Fraud with customer loyalty programs is like barnacles clinging to the bottom of a sailboat: while hardly visible on the surface, they significantly slow down the course. Your program development costs could have paid off in 2-3 years, but instead, they don't even after 5 years.

Popular fraud schemes

"Ghosts" and fake clicks. Let's start with the simplest and relatively harmless scheme—fake lead engagement by unscrupulous lead generation contractors. You pay money to advertise your chain and your program, you get new customers in your store, but sales are still stagnant. Why is that? Well, all of these new clients have been generated in bulk, either manually or by bots, just to get paid for ostensibly expanding your client base. Sounds like a contemporary version of Gogol's Dead Souls, a Russian classic. Click fraud has the same effect of depleting your advertising budget without a tangible return. y
These schemes are relatively benign because they may cost you money but not your reputation. Some fraudsters can inflict much more harm by meddling directly with customer rewards. They would place orders using fake accounts. What does it look like?
Welcome bonuses. Let's say you've decided to onboard new customers by providing nice bonuses or promo codes for registering on your website. And that's where scammers come in. They create numerous accounts with unique phone numbers and then advertise on  relevant websites. The only catch is that the product must be bought through their account: "You get a discount, we keep the reward."

The customer receives the goods at a discount and is quite happy with it. The scammer receives either a cashback for the purchase or additional bonus point to be used at his discretion. What is the bottom line? You register a one-off purchase, but the rewards program does not bring in a loyal customer, and the discount you made is largely wasted. 

Account takeover.  This is the most disgusting and dangerous type of fraud: in addition to financial damage, it also harms the reputation of the online retailer.  

Tools such as brute-force attacks, malware, phishing based on social engineering, leaked logins and passwords from other websites (it's no secret that people often use the same credentials for various online stores) may be used to hack real accounts with rewards. What happens next? Such accounts are resold on relevant forums (not only on the darknet). For example, an account with 10 thousand bonus points can be bought for $20. These points are used to purchase, say, electronics, which is then resold at bargain prices.

Apart from sheer theft, these schemes involve reputational damage to the retailer. Of course, a customer losing a substantial number of honestly earned bonus points will feel upset about not only the scammers but also the online store. In the end, instead of acquiring new and possibly loyal customers, the business loses the ones it already has. 

Unpredictable website overload. Fraudsters are especially active during major sales, discounts, and promotions. A presumably flawless system collapses, customers are leaving, and the discounts fail to yield the expected result.

What can be done?

Artificial intelligence and machine learning can help identify multiple fraudulent accounts or detect account takeover. The system learns to recognize users by their online behavior, from analysis of the device and its environment to the mouse and keyboard usage patterns, etc. This way, legitimate users are separated from fraudsters. 

As a result, loyal customers enjoy solid protection of their accounts and bonus points, while promotional marketing and advertising become way more meaningful. 


we recommend
What do we mean when we talk about hybrid cloud?

What do we mean when we talk about hybrid cloud?

Cyber resilience: business continuity and disaster recovery as default

Cyber resilience: business continuity and disaster recovery as default

Powering exceptional customer experience by data-driven, intelligent retail

Powering exceptional customer experience by data-driven, intelligent retail

Multiphysics Modeling Provides Insights for Commercial and Industrial Solutions

Multiphysics Modeling Provides Insights for Commercial and Industrial Solutions

We use cookies Cookie

Продолжая использовать данный веб-сайт, вы соглашаетесь с тем, что группа компаний Softline может использовать файлы «cookie» в целях хранения ваших учетных данных, параметров и предпочтений, оптимизации работы веб-сайта.