It's not clear to me that communication and use of email addresses works significantly better than communication and use of phone numbers. I don't have any data, but I would say in my experience if you read someone a phone number it is more likely to be transcribed accurately than if you read them an email address. That's probably because there is less scope for mishearing the names of digits than the names of English letters; and phonetic alphabets have never caught on among the general population. The other issue here is standards. Until the mobile phone era, different countries used different mappings of letters to numbers when using letters in phone "numbers". That's why the UK abandoned them (in the 1970s, I believe). Through most of the 80s and 90s, when US companies published alphabetic phone "numbers", non-US callers had to perform a convoluted translation back to digits. Only in the US is international traffic so insignificant that such a situation is tolerable. Paradoxically, though, the UK then adopted alphanumeric postal codes while the US uses numeric codes. In all of this, if you are claiming usability benefits, you need to say what your metrics are. Ability to recall numbers? Ability to dial numbers accurately? Ability to communicate numbers accurately, e.g. looking them up in a phone book? Efficiency or effectiveness? You also need to define the users you are targeting: the average, the old, the disabled? Michael Kay Saxonica |