I was amazed by the report  that Wells Fargo has ended up paying $16m for failing to provide phone services for deaf customers.

Part of the amazement is that in the UK phone services for hearing impaired customers are well established and well used by large companies. In the UK, nearly all banks offer a textphone service and there are well established providers such as TextRelay that offer both operator assist (textphone to voice or vice-versa) direct textphone to textphone communications.  

Wells Fargo seems not to have offered such a service due to security grounds, but this seems to me a misunderstanding of where telephony channel security has strengths and weaknesses, and what conditions you can legitimately attach to a service you offer. 

The text-phone based services for the deaf are as secure (or insecure!) as voice telephony. For the majority of banks today, authentication is based on ID, password or code words;  all of which only prove what you know, not who you are.  Whether this verification is done in a web browser, an IVR or by a text-phone, all are still only authenticating knowledge. If voice biometrics are used, then there can be a significant element of authenticating the who the is person but, as I’ve blogged before (see: “Voice biometric security after the UK DWP ends trials”), the use of voice biometrics is not a panacea for authenticating external callers or an absolute method of proving identity.

The problem for Wells Fargo, though, seems to have been relay calls with an operator rather than the standard text-phone. A relay call is where an operator acts as the interface between the text medium of the deaf person and the voice medium of the contact centre agent.  Generally in the UK this problem is avoided by having a dedicated text-phone team in the bank’s call centre to receive the call (as you can see in these examples of Barclays or Lloyds TSB). This removes the need for an operator and means that there are no extra security issues.  It is a surprise that Wells Fargo didn’t adopt a similar model (or make textphone more public), as the number of customers with impaired hearing can easily justify a dedicated team within a contact centre.  As an illustration, about 450,000 people in the UK use text-phone and around 9m have sufficiently impaired hearing to need assistance . The figures from the US Census Bureau suggest between 1m and 8m Americans have a significant hearing impairment, depending on how you measure severity, and a substantially greater number have some level of hearing impairment that requires assistance.

For those with significant hearing impairment who don’t use textphone, an operator service is usually the only way they can conduct complex interaction by phone. The obvious problem with leaving the operator in the call is that the operator will have potentially have access to all the confidential information that both parties provide. There are ways to manage this, for example insisting that text-phone providers security check their operators, much as a bank has to already vet its contact centre agents. This may not eliminate fraud, but at least means that the likelihood of it happening should be substantially reduced.

Of course, in future the text-phone may be made redundant by video. At the moment, most video trials in retail banking are taking place in branches for the remote advisor scenario. This is where a customer comes to their local branch and is able to have a high definition video interaction with an expert who could be located in any other branch or part of the bank. This the type of technology Bank of America is piloting (and I’ve blogged about in my last post on Bank of Moscow), but its implications are far beyond the branch.

In the UK there is a well established video contact centre for the deaf, SignVideo, which has been running since 2004. This service provides a sign language operator via video link for hearing impaired users.  The focus to date has primarily been on making local government and health services available to the hearing impaired, but it is easy to see how such a model of a video contact centre with a sign language trained advisor could be used by a bank to offer sign language as a national service to its branches. Video would make cost effective what previously would have been impractical.

Imagining video service between points with a bank leads logically onto the idea of offering video externally. Where there is a text-phone, even today Skype can offer enough screen resolution to make sign language practical.  For the UK banks this will probably only represent a technology and security challenge, as most UK banks already have a dedicated team supporting text-phones. For other banks it may be substantially harder to create the operating processes, but video may well provide the impetus.

It would be an interesting development if video provides a cheaper and more accessible way of offering banking, as that tends not to be the perception of the technology or the perception of what many banks are aiming to do.