Updated Voice Biometrics Market Forecast through 2014

Yesterday, I posted projections for the global voice biometrics market (2009-2011). Derek Top, of Opus Research, was kind enough to send me a revised forecast covering the years 2008-2014.

As you can see from the chart, the market is expected to grow from $124 million (2009) to $260 million (2014). This represents a compounded annual growth rate of approximately 16%. According to the chart, the steepest growth is expected to occur between 2010-2011. So let's make this the year of voice biometrics!

Voice Biometrics Basics – Metrics

There are several metrics used to measure the success of a voice biometrics system:

• Enrollment rate – % of authorized users who successfully enrolled (i.e., “base” voiceprint captured)
• Verification rate - % of authorized users who are successfully verified (i.e., “sample” voice print captured and matched against “base” voiceprint)
• False acceptance rate (FAR) - % of unauthorized users who are accepted by the system. Also known as false positive or type I error.
• False rejection rate (FRR) - % of authorized users who are rejected by the system. Also known as false negative or type II error.
• Crossover error rate (CER) – point of intersection when FAR and FRR are plotted against each other (i.e., FAR = FRR). In general, the lower the CER the better.

It’s a fine art to tune a voice biometrics system to achieve simultaneously low FARs (i.e., keeping the unauthorized folks out) and low FRRs (letting the authorized folks in).

Voice Biometrics Basics – Verification

Verification is the process of comparing a person’s “sample” voiceprint to their “base” voiceprint obtained during the enrollment process. If the voiceprints match, based on a probability score, the person’s identity is verified.

Text-dependent voice biometrics systems require a person to speak the same pass phrase(s) during enrollment and verification. Text–independent systems do not. The person can be asked to repeat a series of random pass phrases to capture a “sample” voiceprint. Or, the “base” and “sample” voiceprints can be obtained in the background – even without the person’s knowledge (ideal for forensic applications).

Automated Speaker Verification (Voice Biometrics) - What's Everyone Waiting For?

Automated speaker verification uses voice biometrics to verify a caller’s identity by his or her spoken voice. Voice biometrics measures the person’s vocal tract characteristics (to produce a digitized voiceprint), and is unique to each individual. Voice biometrics is highly accurate, non-invasive and ideal for identity verification over the phone.

Automated speaker verification consists of two processes -- enrollment and verification.

Enrollment:
• Person enrolls (after their identity is verified) by repeating a pass phrase three times (text-dependent approach)

Verification:
• In subsequent calls, enrolled person is prompted by IVR to repeat their pass phrase
• Successful verification allows person to proceed in IVR or transfer to an agent (who is alerted to person’s verification status)

There are four compelling reasons why automated speaker verification should be deployed at call centers:

• Reduces costs
• Increases security
• Enhances customer experience
• Facilitates regulatory compliance

Reduces Costs

In today’s environment, most call centers ask callers a series of challenge questions (e.g., “mother’s maiden name”) to verify their identity. This process typically takes anywhere from 30-60 seconds. Automated speaker verification can do it in less than 5 seconds. Multiply the savings by 100s/1,000s of callers and the productivity gains start to add up. Of course, automating the identity verification process also allows agents to spend more valuable time actually servicing callers.

Increases Security

Identity fraud continues to be a growing problem for consumers and businesses in the United States. Through the use of social engineering techniques such as “pre-texting” or “phishing,” fraudsters can obtain personal information that allows them to assume the identity of another person. This is of particular concern to call centers, as a caller may know the answers to the challenge questions but in fact be an imposter. Automated speaker verification can significantly reduce this threat by ensuring that the caller is actually who they say they are.

Enhances Customer Experience

“Please verify your address.” “What was the name of your first dog?” “Who did you date in fifth grade?”

Automated speaker verification provides a quicker and more user-friendly process for verifying a caller’s identity. For starters, there is no need for the caller to remember or disclose personal information in response to all those annoying challenge questions. Callers get to the service or information they need quicker and spend less time on the call – and that’s a good thing!

Facilitates Regulatory Compliance

Regulations such as FFIEC and HIPAA mandate higher levels of identity verification, known as multi-factor authentication. The three primary factor categories are “something the caller knows” (e.g., account number), “something the caller has” (e.g., hardware token) and “something the caller is” (e.g., person’s voice characteristics). Call centers can use automated speaker verification, combined with another factor (such as account number), to ensure compliance with the multi-factor requirement.

Conclusion

Automated speaker verification is a proven technology which can reduce costs, increase security, enhance the customer experience and facilitate regulatory compliance.

What's everyone waiting for?

Three Good Reasons Why Call Centers Should Use Automated Speaker Verification

Call centers should all be using automated speaker verification (voice biometrics) in conjunction with their IVR systems. Here are three reasons why:

1) May reduce operating costs (e.g., eliminates the need for agent-led, manual verification of callers -- savings of 20-40 seconds per call)
2) Improves customer service (gets the caller to the service or information they need quicker)
3) Enhances security (reduces potential fraud due to social engineering)

Also, financial institutions and health care providers can satisfy multi-factor compliance by using voice biometrics.

Measuring the Accuracy of Voice Biometrics

One question that frequently arises is the accuracy of voice biometrics. Organizations want to ensure that the voice biometrics application will prevent the bad guys from getting in, while at the same time, ensuring that the good guys can get in.

Basically, voice biometrics accuracy is measured by two factors:
1) False acceptance rate (FAR) – unauthorized person is accepted
2) False rejection rate (FRR) – authorized person is rejected

FAR and FRR can be plotted against each other. The point of intersection is known as the crossover error rate (CER). It should be obvious that the lower the CER the better.

It’s a fine art to tune voice biometrics applications to achieve simultaneously low FARs (i.e., keeping the unauthorized folks out) and low FRRs (letting the authorized folks in).