DAVID HICKSON's SILENT CALLS BULLETIN

June 2006

Unreliable Answering Machine Detection

The "unrecognised" cause of Silent Calls

Ofcom uses descriptions and definitions for the term "Silent Calls" that differ widely. Given the true nature of the persistent misuse powers, this should not present any problem,

Does it matter If Ofcom recognises all causes of Silent Calls?

The nature of its powers requires that Ofcom make a determination of persistent misuse in each case that comes to its attention. Any use of a telecommunications network or service that represents a pattern of behaviour or practice likely to cause unnecessary annoyance, inconvenience or anxiety may be subject to a section 128 Notification specifying a particular activity by a particular "person" to be persistent misuse. Following such a Notification, Ofcom may take enforcement action to cause the activity to be brought to an end and not repeated and for the consequences to be remedied, and it may also impose a financial penalty.

Ofcom is required to publish a Statement of Policy covering use of the powers, which it must "have regard to" when using them. This Statement does not however need to specify all circumstances that may give rise to use of the powers.

 

Despite the fact that it is required by law to use its persistent misuse powers as described above, Ofcom pretends, and proceeds as if, it is able to use them in a manner similar to other powers that it holds to cover other responsibilities under different legislative provisions. Ofcom declares that it has introduced "New rules to protect consumers from silent calls", which represent "an enforceable framework for the responsible use of automated calling systems". Ofcom states that it will "take action to enforce the new rules".

My latest complaint to Ofcom about a Silent Call

On 20 June 2006 at 14:15, I received a Silent Call made by LBM, a major UK call centre operator, noted for its shirt sponsorship of Lancashire County Cricket Club.

An investigation by LBM confirmed that this Silent Call was made, and I was informed of the circumstances that caused it. Equipment used by LBM mistakenly detected that the call had been answered by an answering machine.

Many users of automated calling systems, including LBM, use "Answering Machine Detection equipment" (AMD). This equipment analyses the announcement given when the call is answered and attempts to determine whether the sound is "live" or recorded. It is widely recognised that all examples of this type of equipment are unreliable and commonly give false results.

Call centre operators may know how often AMD fails to detect that a call has been answered by a machine, as the agent may record the fact. To my knowledge, nobody has conducted any research to try to find out how often AMD fails to detect that a call has been answered by a person.

Those of us who receive Silent Calls have no way of knowing whether the call is terminated in silence because of failed AMD, or for some other reason. Ofcom appears to assume that all reports of Silent Calls are of calls that are "abandoned" because no agent is available to handle the call. Records of the number of "abandoned" calls are retained and would be used by an Ofcom investigation. There is no record of how many Silent Calls are caused by failure of unreliable AMD, because the equipment cannot know that it has made a mistake.

I have made a complaint to Ofcom about this call from LBM, however

The problem with this case and other similar cases

* Making Silent Calls using AMD is not a breach of Ofcom's "rules to protect consumers from silent calls"

* Ofcom apparently regards using unreliable AMD that causes Silent Calls as "responsible use of automated calling systems"

* As Ofcom will only "take action to enforce the new rules" I cannot expect my complaint to be investigated

* Ofcom appears to assume that all reports of Silent Calls relate to what it carefully defines as "abandoned calls". It is thereby likely to be misled in any investigation

I have been urging Ofcom to consider this issue properly for some time and have previously reported a similar case.

Many users of automated calling systems, including LBM, are now belatedly starting to use the Informative Message to avoid "abandoned calls" resulting in silence. They are also providing CLI to facilitate detection of their misuse, although sadly not for its proper purpose. From cases involving those who are using the Informative Message, we may now start to get some idea about just how many Silent Calls have always been caused by AMD failures.

Ofcom does not intend to review its "rules" until sometime in 2007/8.

The Direct Marketing Association

In its "Brookmead Report", the DMA identified what it called the "Silent Call Chasm". This is the gap between the number of Silent Calls reported as being received, extrapolated from survey data, and the number likely to have resulted from "abandoned calls", based on known automated calling activity. The worst interpretation of rather vague figures suggests that as little as 10% of the Silent Calls received were caused by "abandoned calls".

Although the Brookmead Report noted that AMD was a cause of Silent Calls, it made no attempt to quantify this, or even to suggest it as part of the explanation for the "Silent Call Chasm". In its response to the Brookmead Report, the DMA has so far taken no action in respect of its members, including LBM, using AMD to make Silent Calls.

I have been informed that the DM Authority (part of the DMA) will be conducting an investigation, after seeing a copy of my complaint to Ofcom.

Points for Ofcom to consider

All of Ofcom's "rules" are focussed on "Abandoned calls". They even cover abandoned calls that do not result in silence. Ofcom does not formally acknowledge use of AMD as a cause of Silent Calls. Could Ofcom have missed 90% of the Silent Calls problem?

Anyone who intends to make an interactive voice telephone call would benefit from a means of avoiding the waste of money and time caused by being connected instead to a machine. Sadly, there are no regulations requiring those who do not answer in person to give a clear indication of this fact to the caller, either in advance or at the time of answering.

As automatic answering devices no longer include the detectable clicks and whirrs of mechanical tape machines, I understand there is no reliable means of AMD available.

My suggestions for moving forward with this issue

If anyone were able to prove that they are using AMD that is reliable, there would be no reason to prohibit its use. The burden of proof must however fall on the intending user.

I can see no reason to "permit" use of unreliable AMD as it causes Silent Calls in two ways:

1. A mistaken failure to detect that a person has answered the call causes a Silent Call.

2. Its use introduces a delay into the process of connecting an agent after a call has been answered. This causes many recipients to hang up immediately on the (potentially false, but not entirely reasonable) assumption that this short pause indicates a totally silent call. It is worth noting that this example of a Silent Call is specifically permitted by Ofcom's rules, and that such calls would not be counted as "abandoned calls" in an Ofcom investigation.

I have heard that one example of AMD (which could be typical) would claim to offer reliable results on a sound sample of 10 seconds duration. It would clearly be unacceptable for a caller to wait 10 seconds before deciding how to handle the call after it had been answered. Detection of repeated phrases such as "Hello! Hello!! is there anyone there?" would be likely to facilitate detection of a person.

 

Ofcom has no powers by which it could directly "prohibit" use of AMD. Evidence of use of unreliable AMD by any person would have to lead to the issuing of a s128 Notification of Persistent Misuse. This evidence would provide Ofcom with reasonable grounds for believing that someone's intention to use this equipment would represent a pattern of behaviour or practice likely to cause unnecessary annoyance, inconvenience or anxiety. If Ofcom wished to make this legally enforceable then a s129 Enforcement Notification would have to be issued also.

If Ofcom wishes to have directly enforceable regulatory powers to impose and enforce general requirements on users of telecommunications networks and services then it must approach parliament, or ask the DTI to do so on its behalf. I have already suggested that the Trade and Industry Select Committee may wish to look into the way in which Ofcom is using the persistent misuse powers that exist.

My contact details

You can contact me for advice, information or anything else at:

Silent {dot} Calls @ ntlworld {dot} com