Contact centre

How to make voice a frictionless channel for customer comms – without letting the fraudsters in


Preventing fraud is crucial for contact centres, but poor authentication procedures are damaging customer relationships. What can be done?

29th Jun 2022

Even in the digital age with email and online chat a few clicks away, voice is still one of the most important channels for building and maintaining strong customer relationships. It provides a much-needed human touch that is often absent from digital channels and is useful for dealing with a variety of customer needs, particularly more complex or urgent requests.

However, for all its value, the voice channel has long struggled with the balancing act between customer service and security. Telephony needs to provide an approachable and frictionless experience for legitimate callers, while also protecting them from fraudsters attempting to access their accounts.

Recent Call Centre Management Association (CCMA) research delved into the current state of play for voice. Supported by Pindrop, the CCMA research aim was to highlight the greatest challenges standing in the way of providing a smooth customer experience – without opening the door for criminals.

Poor authentication in voice channels is damaging customer relationships

Call services dealing with customers’ accounts are faced with employing authentication and anti-fraud processes designed to prevent imposters from stealing personal details or accessing accounts.

This has traditionally been carried out by knowledge-based authentication (KBA), a set of questions that usually include personal details and some form of password or PIN. This process is often cumbersome and tedious for the caller, with the nature of telephony meaning it can take several minutes, compared to the seconds required for online verification.

Accordingly, approximately a quarter of respondents to the CCMA research aged between 18-34 (26%) and 35-54 (25%) said they felt accessing accounts over voice could be more straightforward.

Compounding this, by necessity this ‘digital frisk’ usually needs to be completed as soon as the call starts, which can deliver an unpleasant first impression and takes the focus away from the customer need.

Authentication is off-putting at the best of times, but in the worst-case scenario it can actively cause customers to take their business elsewhere. Over a quarter of respondents told the CCMA that authentication issues had caused them to leave a provider. The number increased to nearly half for respondents aged under 34, showing that younger customers generally have a lower tolerance for poor authentication processes. Retailers and banks also saw a greater impact than other industries.  

Legacy authentication is failing to safeguard customer accounts

Not only is the legacy KBA approach to caller verification causing friction for legitimate customers, it’s also been shown to fail to protect them from malicious imposters or prevent organisations from suffering fraud loss.

Criminals often target voice as a weak point to gain access to customer data and accounts. About a third (34%) of respondents to the CCMA research said they had experienced an unauthorised access attempt on their accounts through voice in the last year, and over a quarter (26%) had directly suffered losses as a result.

The legacy KBA approach is becoming increasingly ineffective in the digital age due to the huge volumes of data being stolen in data breaches every day. Over 5.1M records are reported to have been stolen in the last year alone.

In fact, further research by Pindrop has found that if anything, fraudsters often have an easier time passing KBAs than legitimate callers do. In a case study with a national contact centre, 92% of fraudsters were able to pass KBAs, while genuine customers only passed 46% of the time.

Bad actors buy and sell stolen records as a commodity on the dark web, giving them easy access to personal information that can be used to complete KBAs. A complete set of financial login data and personal records can easily be bought for less than £50, and highly organised criminal gangs even have their own customer relations teams and refund policies.

The proliferation of personal data available on the dark web is compounded by the fact that people often duplicate their security information. The CCMA’s research found that younger people are more likely to reuse their passwords, with approximately a third (33%) of those aged 18-34 saying they use three or less different passwords across all accounts.

Being hit by a fraudster can have devastating consequences for an individual and is accordingly a huge blow to the organisation responsible for safeguarding their account. An incident can destroy the customer relationship, as well as causing financial and reputational damage if the firm is found to have been negligent in its security duties.

So, if the long-serving KBA approach is no longer enough, what is the answer to protecting customers from voice fraud?

Why AI & ML can hold the answer to fraud prevention 

AI and machine learning (ML)-powered analytics is one of the most effective emerging technologies in the fight against fraud. This technology can rapidly detect signs of fraud using a multitude of factors such as the caller’s voice, device, and call metadata. Even the savviest fraudster armed with all the right KBA data can be quickly thwarted. Crucially, this all happens passively behind the scenes, with no noticeable impact on legitimate customers.

Technology is already making an impact. A report by UK Finance found that losses due to telephone banking fraud fell by 7% in the first six months of 2021, due in large part to the introduction of stronger voice authentication solutions.

The same technology can also be applied to caller authentication, creating a faster and more enjoyable customer experience that reduces the friction and tedium we have come to expect at the start of a call.

In a Pindrop case study, US-based United Community Bank was able to see marked improvements in efficiency and customer satisfaction after introducing AI-powered authentication. The speed of the new process meant the time callers spent on hold went down by an average of one minute and 11 seconds. The saved time led to a 14% increase in calls handled, with abandoned calls decreasing by seven percent. Post-call satisfaction surveys also found a five percent increase in customer satisfaction.

Getting the balance between customer service and security right

Voice remains a powerful customer relations tool, but organisations must ensure their telephony channels can match the customer service – and security – of their online counterparts.

By moving away from cumbersome legacy KBA-centric processes and towards the agility and reliability of AI and ML-assisted authentication, firms can quickly confirm their customers and move on to fulfilling their requests, without wasting additional time on authentication. At the same time, those practiced fraudsters expecting to exploit the voice channel will find their attack path blocked.



Replies (0)

Please login or register to join the discussion.

There are currently no replies, be the first to post a reply.