Overview
The process of identifying (establishing who they are) and then verifying (proving who they are) a caller in a customer service environment is often a cumbersome and time consuming exercise for both the caller and the contact centre.
It usually involves the customer service representative (CSR) getting the caller to answer a series of knowledge questions (such as mother's maiden name, date of birth, recent transactions etc) or to repeat passwords/PINs before starting to answer the real reason for customer calling. The result is more time per call, and therefore more cost per call as well as increased customer wait times.
In addition, the methods used to verify the caller are open to fraud. The knowledge information used to verify a customer is often fairly easy to find out for a determined fraudster. In addition, customers will often select passwords or PINs that are easy to work out and they may also store them in obvious places which are not secure.
Another advantage of identifying callers at the start of the call is that they can be routed to the most appropriate person, agent group or department depending upon who they are, their value to the organisation, recent call behaviour etc. This can be achieved without the need for an agent to perform the call transfer, which is often very frustrating to customers. Whilst the call is being transferred the customer's details can be "popped" onto the agent screen; again improving customer service and reducing call time.
Whilst there have been significant advances in speech recognition technology there is still a question of customer acceptance. Whilst it is accepted that many people are becoming more used to automated systems most organisations will still need to be sympathetic to their own customer requirements and concerns. They can achieve this through involving customers in defining initial requirements and usability testing as well as marketing the new process in a way that sells the benefits and answers concerns.
Interestingly, recent advances in technology in the area of biometrics mean that callers can now be identified and verified based on characteristics their unique vocal patterns. This adds an additional level of caller convenience and as well as increasing security.
Current Methods Of Automating The Caller ID&V Process
It is important to understand that there are usually two sets of customers that need to be considered. In each case the process may be the same but the customer experience is likely to be different.
Customers using a non-automated channel such as a contact centre.
Customers subscribing to an automated IVR service.
DTMF Vs Speech – both these are appropriate means of data input. Each has its own advantages and disadvantages and therefore careful thought should be given as to which to use. For example, using DTMF on a mobile telephone is possible but cumbersome whilst some people may prefer not to speak personal details out aloud. DTMF is at best restricted to inputting numeric data whilst speech recognition opens up a much broader vocabulary.
Caller Identification - How Does The Caller Identify Themselves?
The first question to ask is how does the customer currently identify themselves and is it possible to use this as a "token" in the automated system (as this is more convenient for the customer).
Typically, a numeric identifier such as account number or membership number can be used. This is usually easy for the customer to remember or find. If there is no obvious candidate then a token has to be chosen but bear in mind that numbers are easier for the system to recognise and should be easy to remember therefore customer self-selection may be required.
Note: Whilst callers could be identified from their CLI (Calling Line Identity) there are inherent risks attached to this in that a stranger could be using the phone. However, CLI can be used to add additional security in certain call handling scenarios.
Caller Verification - How Does The Caller Prove They Are Who They Claim To Be?
Again, ask what questions/answers are currently used. These may or may not be suitable as tokens in an automated system.
Typically, answers to knowledge questions such as mother's maiden name are not suitable (due partly to recognition accuracy but also how to create the grammar to recognise them in the first place and what other questions could be used if that answer is compromised).
Most solutions revolve around the introduction of PINs or new passwords. Callers are asked to speak back (or use DTMF) a particular digit from the PIN or the location of a particular letter in their password.
Failed Identification Or Verification
Callers failing ID&V (or as a result of unusual call activity) will be transferred to a CSR who can carry out the task or help the caller with their ID&V problem and transfer them back into the automated system.
Return On Investment (ROI)
Typically, anything up to half of each call's duration can be spent on identifying and verifying the caller. Using a CSR to perform this task is expensive compared to the cost of automating it. CSR's can instead concentrate on adding value to each call. Additionally, because the automated process is quicker it will require less incoming lines and telephony charges will be less.
Screen popping caller's details on a CSR's screen also saves time and improves customer satisfaction. Slightly harder to quantify are the benefits from routing calls more effectively as this is a judgement on customer satisfaction. Likewise, the benefits of reducing fraud are difficult to quantify but no less obvious.
Evolution
Whilst the benefits of automating the ID&V process are compelling in their own right new advances in technology mean there are even more reasons to adopt automated ID&V.
Speaker Verification – Voice Biometric?
Using unique characteristics of a person's voice a voiceprint (or signature) can be created for each customer. When they subsequently call in their utterances can be compared to the stored voiceprint to confirm their claimed identity. Customers don't even need to remember a password as they can be asked to repeat random digit strings.
The advantages are greatly improved customer convenience and satisfaction (customers are overloaded with the number of passwords and PIN codes they need to remember) as well as a significant improvement in security.
Summary
The manual procedures for identifying and verifying customers used in many of today's contact centers are unsatisfactory both in terms of customers convenience and increasingly from a cost view point. Coupled with this, the significant improvements to security that can be achieved by deploying the latest speaker verification technologies mean there is a compelling case for considering the automation of the ID&V process.
Why Choose ICR?
ICR has real-world experience of developing and delivering automated ID&V solutions encompassing the latest technology. ICR is independent from technology vendors and are therefore able to provide truly objective advice and provide technical solutions utilising best of breed technologies.
About Connect Managed Services:
Connect are the independent communications providers who can transform how your organisation communicates – both internally and externally. We deliver solutions and services that join up employee and customer communications across platforms, across sites and across countries, in three core areas: Contact Centre Solutions, Unified Communications and Network Services. We provide simple, elegant solutions to the most complex problems.
Published: Tuesday, June 8, 2004
I am checking out all the amazing and daily updated content on ContactCenterWorld.com and networking with professionals worldwide
Send To Friends Post On My Wall