• Tailored self-help pages
Self-help pages can be automatically tailored for
individual users, based on the types of application
they run, making the relevant information more
readily available than is generally the case today.
• Monitoring of newsgroups
Public-domain newsgroups already include a wealth
of information from users who have had issues and
potential fixes. Information from these news groups
can be used to:
— populate self-help pages,
— develop diagnostic tools,
— provide call centre agents with more information
to diagnose the faults.
Although there is likely to be significant value in
doing this manually, an even better approach would
be to use AI tools to parse the newsgroups and
automatically extract the relevant information.
• Broadcasting known issues
As help desk calls are handled, common issues and
faults will occur. By using ‘push’ technology to
make this information available to users who might
become affected, customer problems, and hence
support calls, can be reduced.
4.3 Receive problem report and establish
ownership
Determining where the problem lies in a multi-provider
environment is difficult. Several ways of addressing this
issue are discussed in the following sections.
Making use of all available data
A user’s description of the problem, although important
to obtain, can be misleading. Users frequently give
incorrect information for reasons such as:
• they do not want to get charged for help,
• they do not understand the question and give an
arbitrary answer so they do not appear
unknowledgeable,
• they think they understand the question but do not,
and hence give an incorrect answer.
There are several additional data sources potentially
available to support providers, such as logs held on
home hubs and DSLAM information, that can be used
to supplement the information obtained from the user.
The key issues are how to gain access to the relevant
data at the time the support contact is made, and how
to interpret it quickly and accurately.
There are also issues with organisational boundaries.
For example, the ADSL wholesale provider owns
the DSLAM data whereas the ISP and/or customer own
the hub and associated data. Comparing these two sets
of data can be helpful in diagnosis, but ensuring the
support organisation has access to both data sources
can have significant regulatory and technical implications.
Electronic interfaces to other organisations
By providing electronic interfaces across organisational
boundaries, a whole host of problems can be addressed.
For example, a user could call an IPTV provider to ask
why their video stopped half way through. The IPTV
provider could determine from the customer’s ISP that
the 1 Gb download limit had been reached. Taking this
scenario one stage further, the IPTV provider could
increase the download limit for the customer, acting as
the agent for the service provider and receiving
commission for doing so. There are industry standards
being proposed to enable electronic interfaces between
organisations to be set up more easily [1].
4.4 Allocate problem to the correct team
Once the organisation with responsibility for resolving
the customer’s problem has been identified, it is then
necessary to identify the root cause of that problem so
that it can be allocated to somebody in the correct
geographical area and with the necessary skills to fix it.
The first key enabler to this is test and diagnostic
tools that can correctly identify the skills necessary to
resolve the problem, even though at this stage the
precise cause of the problem remains unknown.
For example, if it is possible to determine that the
problem is in the IP layer, then we need to allocate it to
somebody with IP diagnostic and configuration skills.
For example, within BT Wholesale, a sophisticated
tool has been developed that tests all layers of the
broadband service, from the copper bearer through
xDSL and ATM layers to IP and the service running over
IP. It does this in a single ‘one-shot’ check, and
combines the results of the tests at the various layers in
order to identify what skills are most likely to be
required in order to resolve the problem.
The second key enabler is how work is allocated to
engineers. By keeping a detailed database showing
which engineers are working, the skills that they have,
and their geographical location (if relevant), the most
appropriate person can be found to work on the
customer’s problem. A sophisticated work allocation
system has been developed within BT in order to do this,
which is the subject of a paper by Li [2].
Six steps to excellent customer service
BT Technology Journal • Vol 24 No 1 • January 2006 125
4.5 Establish root cause of problem and resolve it
At this stage, detailed diagnostics take place by the
person allocated to resolving the problem to identify the
root cause, and (usually) the same person also effects
the repair.
This can be a very time consuming and expensive
business, and is an area where the application of new
technology can deliver significant benefits.
Enhanced test tools
A new family of test tools is now becoming available,
often based around interaction with a field engineer’s
laptop computer or PDA. These have a great deal more
intelligence built into them, as well as new test
techniques such as those based on using xDSL
transmission parameters. One example of such a device
is the CopperMax/OSP Field Tester, jointly developed by
Spirent and BT [3].
IP-layer diagnostic tools
IP diagnostics is still a developing area. Much reliance is
usually placed on basic IP diagnostic utilities such as
Ping and Traceroute. However, the effective use of such
tools requires high levels of skill, which is unrealistic for
a mass-market product. The inherent unpredictability of
an IP network also makes it difficult to draw definitive
conclusions from such tests.
Some companies are now developing easier-to-use
IP diagnostic tools that are aimed at mass-market or
‘telco-scale’ users. However, the widespread adoption
of such tools has not really started.
Diagnostic agents
Software agents running at strategic points on the
network, e.g. on a home gateway, can be of great
assistance when trying to find the root cause of a
problem. These may, for example, inspect the
configuration and performance of a user’s xDSL
modem. Another form of such an agent would be
software that can inspect a detailed activity log
associated with a particular service, such as e-mail.
Again, the use of such agents is not yet widespread,
although they could become highly effective tools for
root cause diagnostics of broadband problems.
Self-healing networks
In some cases the entire ‘resolve problem’ activity can
be avoided by using self-healing techniques, whereby an
identified problem is fixed automatically. For example,
duplicate DHCP servers could be handled in this way.
Full automation of self-healing does, however, throw
up many issues. In the case of the duplicate DHCP
server, how would the support provider know which one
should be switched off? Also, how is the user kept
informed, if at all?
4.6 Inform the customer that the problem has
been resolved
When the service provider thinks the problem has been
resolved, they should ideally confirm with the end user
that they agree. Some of the self-help capabilities
discussed in section 3.2 can be used by the customer for
this purpose.
The effectiveness of this step is greatly enhanced if it
can be performed shortly after the ‘resolve problem’
activity, ideally by the same person. In this way, if the
customer does not agree that the problem has gone
away, further attempts can be made to resolve it, for
example before an engineer leaves the relevant
location.
The key issues here are:
• ensuring that the customer can be contacted, by
ensuring relevant information is kept on the CRM
system and making this information readily
available to the person who is charged with
resolving the problem, and
• providing a quick, easy and reliable method that
the customer can use to verify for themselves that
their problem has been resolved.
4.7 Reducing support volume
Interoperability issues
Understanding interoperability issues between equipment
from different vendors can greatly increase the
chances of a customer’s service working correctly first
time.
This requires that equipment vendors, service
providers and other relevant parties co-operate. For this
to happen, there must be incentives for all involved in
the form of increased sales and reduced support costs.
There are some industry bodies that have been set
up to tackle this issue. This is discussed in more detail in
Bull and Harrison [4]. However, this area is still in its
infancy, and the potential for greater co-operation and,
hence, a significant reduction in demand for support
through increased industry collaboration, is significant.
Also, by providing known combinations of equipment
that work, the number of ‘I cannot get it to work’
calls will be reduced. This could be achieved through
bundling of products that have been tested together
and are known to be compatible, either into single
Six steps to excellent customer service
126 BT Technology Journal • Vol 24 No 1 • January 2006
‘service packages’ made available to customers, or
through a publicly available ‘compatible equipment’
list, similar to those published by computer equipment
vendors who have tested various releases of operating
systems on specific hardware platforms (an example of
this can be seen on Hewlett Packard’s Web site [5]).
Analysing trends in support requests
Analysing trends in support requests can also help to
reduce demand on the support service. Careful
monitoring of problem reports, especially those
received shortly after the product has been launched,
can result in changes to the equipment or service
design, and its supporting literature, that will reduce the
number of support requests.
Common ease-of-use issues can also be identified
and then addre