For businesses there are many ways they can transform customer service, providing a friction-less experience over IVR is one of the most important metrics to be analysed. This is because for any caller the first interaction they have with a business is through an IVR. Therefore, it is ideal that when someone calls the interactive voice response (IVR) technology must be smart enough to identify the caller. There are many ways to make it possible.
Identification & verification (IDV) : In the IVR ,authentication is done by IDV. It is based on correct ID of the caller by which not only caller's name is revealed but also their account histories come into picture. A single customer profile is maintained for every customer and thereby it helps to serve them more quickly and effectively.
Another aspect of IDV is the more confirmation about the customer, the better trust they have on the organisation. Keeping this is in mind there are few IVR authentication best practices which help to identify whoever is on the other side of line.
There is a common practice of asking customer name at a first interaction but it is far more impressive to know the caller even before picking up the call. This is possible by ANI where a call comes and ANI matches the caller's number and shows the customer profile in details like their location or the name of organisation from where they are making call.
In today's world a house owns multiple phone numbers ( landline/mobile) . ANI database allows to add new or additional phone number to each account that increases verification efficiency.
Many organisation need second layer of authentication . Before going for it client must agree to provide extra piece of information such as birth date, social security number etc. Mostly last four digits of social security or similar personal id is used as a second layer of authentication, as many people will hesitate to share their personal information. This type of authentication is only needed to access secure data such as account balance etc.
An IVR can function more deeper than answering a call. It can offer personalized service for each customer by identifying them. For example, The IVR can tell the client about their payment status and that provides the customer an immediate chance to confirm and provide payment without going through lengthy IVR menu. Positive identification of customer reduces the risk of multiple transfer or drop off. If the prediction is right, it is advisable to send text messages in future which will help to inform the customer to pay bill in advance.
Now as the IVR identifies the callers as soon as they call, it is important to know whether the agent needs ID at all. In some industries caller's ID remains private and can’t be passed on to an agent. So the last IVR authentication best practice is to skip authentication.
The first goal of most IVRs is the self-service of IDV. Without knowing the client one can’t provide successful self-service. So increasing IDV will result in a decrease in call center cost and increase in customer satisfaction.
Knowledge based authentication (KBA): KBA is an authentication method that asks questions based on detailed information. This information is usually more advanced than the information that is accessible from email or a computer by a fraudster.
There are three types of KBA
1. Static KBA
2. Dynamic KBA
3. Enhanced KBA
In static KBA , end user pre select questions and provide the correct answer. The question/answer is then stored and used to verify the person's identity later. Static KBA questions are very simple and commonly asked like "What is your pet’s name?" Or "What is your favourite colour?" etc. The problem with static KBA question is if someone share the information on a social media site, the answer can be easily accessible.
Dynamic KBA is the second level of authentication where questions are not generic and only applied to a specific person. It is also known as, "out-of-wallet "questions because the answer can't be found in a person’s wallet if it is lost or stolen. The questions are generated from user’s credit history or public records. It is an attractive option for e-signature users that need to be authenticated instantly.
It is like dynamic KBA where it presents multiple choice questions to users.
Unlike password or token-based authentication, it uses unique biological characteristics to verify an individual’s identity. This is more convenient for the users as no need to remember passwords or carry token that can be easily stolen or lost. Voice recognition also called voice authentication analyses a person’s voice to identify the person. There are few factors like shape and movement of mouth and jaws, airways, soft-tissue cavities influence voice patterns, which creates a unique voice print.
Voice authentication is not the same as speech recognition where speech to text application technology is used. Speech recognition can sense verbal language but it can’t verify identity of the speaker based on unique vocal attributes.
• Widely accessible on mobile phones for authentication as all phones has microphones.
• Cost effective
• Convenient and familiar to users
• Useful for customer service because of phone-based use
• Not as accurate as facial bio metrics.
• Background noise can impact the quality of sample.
• Requires liveness detection to verify whether the sample is a live speaker or a recording.
• Not suitable in noisy or public place.
According to Gartner, “Using voice bio metrics combined with phone printing provides the strongest method for authentication callers and detecting fraudsters.” Phone printing and voice bio metrics offer a multi-layered solution. When good quality audio from contact center is available, voice bio metrics can provide better accuracy. Combining together can identify right caller, right number and right location.
For more assistance on caller authentication and our custom solutions, please write to firstname.lastname@example.org.
Author Bio: Sumita has 12 years of strong experience in various segments of management ranging from contact center to process development with the help of lean Six Sigma methodology. She is an avid reader, occasional writer and an extensive traveler beyond her professional commitments.