• Accuracy - FRR/FAR/FER/EER, speed and maturity of technology
• User acceptance - personal, cultural, political
• Availability of technology
• Competition or lack thereof
• Critical Vulnerabilities
• Standards support - AFIS, BAPI, FIPS, etc.
• Cost - ROI
• Manufacturer reputation and 'history'
• How critical is the data or physical location being secured

The performance of biometric based access systems using fingerprints or other means is limited by the performance of sensors and algorithms. The match confidence is always a matter of probability, and the accuracy is generally measured by establishing the values for the following four criteria: FRR (False Rejection Rate), FAR (False Acceptance Rate), FER (Failure to Enroll Rate), and ERR (Equal Error Rate). These rates are usually expressed as events per 1,000 or 10,000 uses.

FRR (False Rejection Rate) is the rate at which the system incorrectly rejects a legitimate attempt to verify. With increasing FRR, the probability increases that authorized personnel may have to submit credentials repeatedly before being granted the access to which they are entitled. In a high security environment this might not be such an important issue, but a high FRR it may create user acceptance problems when the.

FAR (False Acceptance Rate) - is the rate at which the system incorrectly accepts an invalid verification attempt. FER (Failure to Enroll Rate) - is the percentage of people that do not have sufficient sample quality to enroll on a given biometric system. For applications where convenience and general user acceptance are more important than security, administrators have had to settle for a high FAR in order to assure that authorized individuals are always granted access, at the cost of possibly granting access to unauthorized individuals.

FER (Failure to Enroll Rate) - is the percentage of people that do not have sufficient sample quality to enroll on a given biometric system. Every biometric feature can fail. Examples of failures to enroll can be caused by worn down or unavailable fingertip for fingerprints, medicine intake in iris identification (Atropin), hoarseness or lack of speech in voice recognition, or a disability affecting one's signature. Therefore, every biometric system needs to offer the flexibility of a 'fall-back process' and an alternative method of authentication.

EER (Equal Error Rate) - is the point where the FRR and the FAR are equal. A system with an FRR of 5% and a FAR of 5% will have an EER of 5%. FAR or FRR test parameters are often tuned to favor either FRR or FAR, depending on the value that the test is meant to measure. In order to establish the proper EER value however, the test parameters must remain consistent for FRR and FAR testing. As such, the EER is sometimes considered a more accurate value than separate FRR and FAR values.

User acceptance is often the key for deploying one or the other biometric technology. Regardless of how accurate a system is the more difficult or inconvenient the system is to handle, the more frustrated users will become and the more money will have to be spent on training, maintenance, support and implementation.