The potential role of buyer types for identifying fraud

Marketing data is commonly used to predict when customers will be buying goods or using credit – for example ConsumerView Triggers. What journey does a ‘genuine’ customer go through?

The below diagram shows The McKinsey Consumer Buying Cycle and four main stages of the buying process. This is a generalisation, but illustrates the concept well. It could be considered unusual for a consumer to go direct to ‘moment of purchase’ or to select multiple brands for their final purchase.



But is there a correlation between not needing credit and no intention to repay?

Let’s say we have an affluent member of the public and they apply for a “no credit check current account” or the individual has applied for several £10,000 loans at a high interest rate. That should not make sense because we know, from the data and demographics, that we would not expect that type of person to be applying for those types of product.


We should be able to make the same exact premise and “reverse” marketing data, obviously subject to consent – as we know and can predict who are the people who will be applying for loans, making a credit card balance transfer and so on. Fundamentally these are the types of rules that are also part of fraud systems – for example, “should we refer this application because it’s unusual for the person to have made five loan applications in the space of two hours?”

Furthermore, it is widely understood that there are links between the web navigation journey and fraud, particularly in an e-commerce environment. For instance, when shopping for a new mobile phone a fraudster may go direct to a high value product, add two to the basket and select any tariff before proceeding to check-out. A genuine customer may compare prices and products on multiple sites, research coverage and what’s included in the tariff before leaving the website. They then may return later and contact the seller via other channels before finally deciding to purchase.


In the financial sector, pre-qualification may help to identify fraud in a similar way. Firstly, are genuine customers more likely to use a pre-qualification service (either through an aggregator or with a lender) to maximise their chance of a good deal? Secondly, are fraudsters using pre-qualification services to understand which businesses to target? If so, how can you differentiate between the two?

To help distinguish between genuine customers and fraudsters testing systems, organisations should perform analytics on the source of applications and ensure that they understand the routes fraudsters are taking to target their business. Businesses should also consider taking the same approach to pre-application searches as an insurer would for quotes, who look for inconsistencies and manipulation of the process, same devices with multiple searches and ghost brokerage.