A perspective on security and customer service through voice biometrics is offered by Claire Richardson, pictured, VP – Workforce Optimisation Solutions, EMEA, Verint.
Many have experienced the downside of stringent security and time-consuming protocols in the call centre that have resulted in poor customer experiences—a challenge that has traditionally been unavoidable for organisations wanting to prevent fraud and adhere to data protection laws. From passwords, passcodes and your first pet’s name, customers are asked endless security questions to validate their identities, some of which they may even have forgotten the answers to.
An Opus Research report on the topic of authentication and contact centre security revealed the results from a survey it commissioned “of 1000 individuals who had recently carried out business over the phone”. Results showed that “about 65 per cent found authentication processes to be frustrating” and “nearly 50pc considered the process too time-consuming.”1
Confronting customers with repetitive and time-consuming security hurdles can both slow processes down and run the risk of poor service perceptions among customers. So, how just how can organisations make improvements in this area without compromising security?
It’s all about balance. Fraud detection measures are essential in today’s environments. It’s been estimated that as many as one in 2,500 calls are fraudulent. With such figures, it’s vital that organisations not antagonise the thousands of other legitimate callers in the process of weeding out the impostors.
As new technologies mature, faster and better ways are emerging to help overcome this challenge. Over the phone, the quickest way to identify someone is by their voice, and biometric identification via the unique characteristics of a person’s voice is finally making this an attainable and effective way to streamline the experience and improve customer service offerings. Essentially, when callers start speaking, the technology can match their voices against known “voiceprints” on record to confirm their identities.
There are different motivations for building a voice biometrics capability, most of which centre around two main issues: fraud reduction and customer experience. Depending on the security needs of your organisation, the strategy you choose to implement can be flexible and tie in with existing technology. No modern contact centre should rely on merely one form of identification, and many are likely to already have CRM systems, call monitoring and recording systems, and analytical databases in place. Collectively, these can help in the effort to identify fraudsters and improve the customer service, while complying with industry regulations.
Working with CRM systems, behavioural analysis and voice of the customer analytics, agents can better understand preferences and actions to increase their knowledge of individual customers. This level of in-depth analysis also can help detect unusual behaviours, such as long pauses over the phone or certain intonations of voice, which can signal a potential fraudster.
Further, voice biometrics can help reduce the need for security questions, which are actually a much weaker security measure given fraudster use of social engineering and online ID theft to bypass such questions. And beyond helping combat fraud, agents themselves can benefit from the ability to resolve issues much more efficiently, translating to the customer in the form of faster, better service.
As noted by in the Opus Research report, “Financial institutions can reap significant financial gains by using voiceprints to authenticate through the phone channel.” According to its calculations, “…a large money center bank will see over $100 million in economic benefit from reducing fraud loss in the contact center, lowering operating costs, and providing a more pleasing customer experience.”2
Sources:
[1], 2 Opus Research, Inc., A New Authentication Paradigm: Call Center Security without Compromising Customer Experience, 2013). Report reprinted by Victrio, now Verint Systems, with permission.
Visit http://www.verint.com
