Customer data analytics and intelligence-sharing between companies play an essential role when tracking prolific fraudsters, explains Steve Arnold.
When organisations have the ability to combine and distil raw data into meaningful insight from the sources available to them, they can build a clear picture of fraud-related activity in their contact centres and make quick decisions to combat it.
This value is brought to life when the customer analytics team conducts analysis of the combined contact centre data across all of our customers. And recently we discovered insight that is enabling our banking customers to play a key role in helping fight organised crime, by providing essential evidence of law enforcement groups.
First, a look at the data sources
Large contact centres receive thousands of calls a day, and with each call comes important data about the caller. This data contains valuable intelligence that can help spot prolific fraudsters – you just need to know how and where to look.
In addition to incoming call data, Smartnumbers customers have access to the following three layers of defence when analysing caller risk:
1. Local deny lists – a customer’s own list of blacklisted numbers
2. Consortium data – lists of blacklisted numbers flagged by other customers
3. Machine learning – intelligence gleaned from AI analysis of caller behaviour
By fully utilising a combination of all these sources – and actively sharing fraud insight with other users of the platform – organisations give themselves the best chance of identifying and stopping the fraudsters targeting them and the wider community.
With analysis, we find insight
When we view the combined data for all our customers, we start to see patterns. Here we’ll follow the story of one particular caller.
We could see evidence of the same caller regularly targeting a number of the banks which use our platform. This caller – recognisable by his voice – was found to be using several numbers and switching between them.
Some of our banking customers had the numbers used by this particular caller on their own ‘deny lists’, ensuring calls received from these numbers are instantly being flagged as high risk in their contact centres. Other banks, making use of consortium data, were able to identify that calls from these numbers are considered by other organisations to be high risk and flag them too.
Using machine learning techniques, we could also see this caller was varying their calling patterns, presumably trying to evade detection. We saw them sometimes using the same number over an extended period to target the same bank, before switching to another organisation. At other times we saw them alternating numbers and calling in short, sharp bursts. This kind of information helps build a picture of suspicious behaviour and spot new callers behaving this way.
When suspicious callers are flagged, fraud analysts are able to create and share profiles of fraudsters in the Smartnumbers platform, including details on behaviour and types of attack. In this case, one banking customer found that this fraudster sometimes uses different accents or a voice-changer when speaking to a call agent. And another found that the fraudster worked as part of a team, and could hear other fraudsters operating in the background. This kind of insight can be made visible to other organisations, when logged in the platform.
By logging attack methods, we see how a fraudster is using the telephony channel, which helps our clients anticipate what he’s likely to do next. In the case of this fraudster, he was found to be targeting multiple customers using stolen data, attempting to gather more personal details before initiating ‘account takeover’ attacks at a later date.
The combined intelligence from multiple organisations, logged in a clear, structured format, can provide useful evidence for law enforcement groups. At the time of writing, the banks involved are in discussions about when and how to bring this fraudster to justice.
This kind of insight, supported by in depth analytics, underscores the value of monitoring telephony data and the importance of collaboration to ensure the greatest success identifying fraudsters. It helps our customers continuously refine their fraud management approach and make decisions faster. And it enables Smartnumbers to improve the value we provide to our customers through the platform, strengthening our collective defences against fraud.
