Table of Contents
While many consumers manage routine tasks like password resets, account access, and billing payments through digital channels, many still prefer to call for help.
Authentication & verification over the phone is often a point of frustration. After waiting on hold to speak with an agent, the last thing customers want is to repeat their details before they can address their query.
For your agents, it’s a repetitive and unappreciated task. No one ever leaves a positive review for being correctly authenticated, but customers will quickly leave a negative review if they have to repeat themselves too often.
Automating knowledge-based authentication (KBA) offers a streamlined way to verify identities, ensuring security without sacrificing customer experience.
What is knowledge-based authentication?
KBA is the process of asking callers a series of questions to confirm their identity. It’s commonly used in contact centers, where agents manually verify callers using information stored in their CRMs, such as the caller’s name, date of birth, customer number, or answers to security questions like their mother’s maiden name.
Although it requires a small amount of effort from the customer, KBA is a familiar and generally simple process to follow. It’s considered secure because it’s unlikely that multiple people would have the same combination of personal details.
Not all KBA is the same. There are different types, each with unique methods and purposes:
1. Static KBA
Static KBA uses pre-set questions that users answer during authentication. These are typically personal questions like, “What was the name of your first pet?” Since the user provides the answers upfront, this method can be vulnerable to security risks if someone else gains access to that information.
2. Dynamic KBA
Dynamic KBA generates questions in real-time based on a user’s personal data. The questions vary depending on the available information, making it harder for fraudsters to guess or prepare answers.
How to automate the knowledge-based authentication process
Automating the KBA process over the phone is key to offering personalized service and better self-service options. It also removes routine tasks from your agent’s workload to give them more time to handle more complex and meaningful tasks. With routine processes like authentication managed by an AI-driven voice assistant, your contact center’s efficiency improves, as does your cost-to-serve.
Listening to what the caller is saying
To effectively automate the KBA process, a voice assistant must listen, understand, and respond to what the customer is saying. Listening requires accurate speech recognition to correctly identify a caller by name, number, ZIP code, or other personal information your organization chooses.
Identifying callers with voice AI
Even the best automatic speech recognition (ASR) systems will result in gaps and errors in speech recognition. This means spoken language understanding (SLU) models are also needed to recover important information from incorrect speech transcriptions, using context and customer information to work out the correct input.
Effective AI-driven voice assistants can take alphanumeric strings over the phone and extract names at a phonetic level. Once the speaker’s words have been transcribed, large language models (LLMs) extract meaning from what the customer has said and continue to move the conversation toward an appropriate resolution—in this case, verifying that the customer is who they say they are.
Responding to callers with voice AI
Customers need to feel that their personal information is being handled correctly and securely. After the voice assistant has listened to and understood the caller’s details, it should convert that information into speech to guide the customer through the next steps. Text-to-speech (TTS) models transform text transcriptions into spoken words. Traditional TTS has delivered robotic-sounding voices, but a new generation of TTS can leverage voice cloning and intonation training to deliver a more natural experience when responding to the caller.
Use cases & benefits of knowledge-based authentication
1. Enhanced security with dynamic KBA
Dynamic KBA enhances security by asking personalized questions based on a user’s personal data pulled from a pre-existing database such as your CRM. The questions vary with each authentication attempt, making it harder for unauthorized individuals to prepare answers.
2. Regulatory compliance and fraud prevention
Regulatory compliance is essential in industries like finance and healthcare. Automated systems should have compliance certifications and protocols in place to ensure that every single interaction is private and secure and that every identity verification step meets industry regulations.
3. Cost-effectiveness
Automating the KBA process can significantly reduce operational costs. By handling routine verification tasks through a voice assistant, your contact center can handle call spikes more efficiently and reduce the likelihood of call abandonment. This automation frees your agents to focus on more complex issues and revenue-driving calls while generating lower staffing costs and improved overall efficiency. Customers can easily verify their identity while waiting for the next available agent. Alternatively, they can immediately resolve their issue using a voice assistant with no wait time.
4. Scalability for growing businesses
As your business expands, so does the volume of customer interactions. Automating the KBA process with AI-driven voice assistants means you can scale with your growing needs, ensuring that your verification processes remain effective and secure. Voice assistants can easily adapt to increased call volumes without compromising performance or security. Your contact center can deploy a voice assistant to manage multiple authentication requests at once, allowing your agents to handle more customers seamlessly and maintain high service standards.
Alternatives to knowledge-based authentication
There may be some instances where you want to adopt alternative authentication and verification methods.
KBA may not be necessary for simple requests, such as retrieving basic information or quickly accessing a customer’s support history. In such cases, you might want to consider automatic number identification (ANI).
Automatic number identification
ANI identifies a caller by their phone number and is usually provided by CCaaS (Contact Center as a Service) providers. It matches the caller’s phone number with CRM records to quickly determine their identity, making it easy for callers to start their query without going through a separate identification process.
ANI is useful for enabling your contact center agents to provide personalized service as soon as they answer the phone, but it has limitations. It can fail when customers call from a different number or when multiple customers share the same number.
Voice biometrics
Voice biometrics can be an effective method of identification and verification in managing certain customer requests in a contact center, especially those involving authentication, security, and sensitive information.
This method offers an easy user experience once customers are set up, although the initial setup can be difficult to explain. While the accuracy of identification is medium to low—about 1 in 200 people might be wrongly identified—the accuracy of verification is high, making it hard to impersonate someone, though not entirely foolproof. Implementation can be challenging and costly, as you’ll need to collect voice samples and set up a voiceprint database.
Transform your knowledge-based authentication with PolyAI’s
KBA is a tried and trusted, customer-friendly approach to identifying and verifying customers. The way you choose to identify and verify customers will depend on the level of security necessary to provide customers with access to their accounts and the level of customer experience your company wishes to provide.
PolyAI voice assistants have attained a reliable level of accuracy in performing KBA over the phone. We can identify callers based on any piece(s) of information your company stores in its CRM, including ZIP codes, order numbers, names, addresses, and more.
Conversational voice assistants
Our voice assistants can engage in natural conversation with your customers to automate the KBA process, making it more intuitive and user-friendly. So instead of waiting in a long hold queue for your busy agents to identify and verify each caller, a voice assistant can answer instantly and then either hand off to the right team or complete the transaction.
Data-driven insights
When customers hit snagging points during online verification processes, they will pick up the phone. PolyAI voice assistants can help you identify friction in the customer journey by highlighting peaks in specific queries and issues as they arise. So, if your website’s reset password functionality is not working as expected, our voice assistants can either handle the identification and verification process or give you insights in real time to fix the issue.
Reduced operational costs
In many cases, the KBA process is straightforward, but it can be resource-intensive, requiring significant agent time. By implementing a PolyAI voice assistant, you can automate the low-value part of a call to drive down AHTs by 20-30%, giving additional capacity to your call center without sacrificing customer service. Not only is your team freed up to focus on high-value calls, but you can expect to reduce your seasonal hiring by 40-60%.
Speak to a member of the team today and discover how PolyAI can help resolve over 50% of calls and consistently deliver your best brand experience.
Knowledge-based authentication FAQs
KBA can be used across a range of industries. Some examples include:
Financial services: Banks, credit unions, and other financial institutions often use KBA to verify identity during online transactions or when opening new accounts.
Healthcare: Medical facilities and insurance providers sometimes use KBA for patient verification and protecting sensitive health information.
Telecommunications: Phone companies may use KBA when users need to recover accounts or verify their identity for customer support issues.
KBA questions are chosen based on information that is expected to be known only by the individual being authenticated. There are two main types of KBA:
Static KBA: The questions are predetermined and typically answered when an account is set up.
Dynamic KBA: These questions are generated in real time and more specific to the user. For example, “Which of these addresses have you lived at in the past?”