Tackling Utility Customer Cost to Serve

Utilities in the U.S. stand to save
considerable expense by evaluating
their customer service operations
from a customer cost-to-serve perspective.
They spend on average $42.50
per year servicing each customer, according
to research conducted by The Gartner
Group in 2006. Shaving a mere dollar off
this cost-to-serve figure can amount to
millions of dollars in annual savings for a
large utility.

Utility customer cost-to-serve is the
sum of costs associated with retailing
energy to the customer, including labor
and technology costs, divided by the number
of customers. It’s distinct from other
utility value chain costs like wholesale
energy and energy transmission and distribution.
Utility customer cost-to-serve
generally includes all costs related to
metering, billing, payment, collections
and customer service (see Figure 1).

Activity-Based Costs vs. Departmental Cost Structures

Breaking down the utility customer cost-to-serve into individual components
the first step to savings. This is easier
said than done, however, and most utilities
struggle with accounting cost structures
that do not relate easily to operational

Since utility retail costs are typically
concentrated in the call center and in
operating the large-scale IT systems that
automate billing and payment processes,
it is not uncommon for utility accountants
to break down cost to serve along departmental
lines: 1) call center IT systems and
telecommunications; 2) call center staff;
3) billing staff; 4) IT staff; 5) customer
information and billing IT systems; and
6) outsourced service costs.

This breakdown is relatively easy to
measure and is helpful to those departments
managing cost with “blunt instruments,”
such as headcount reduction.
Unfortunately, it is of limited value in
identifying cost-saving opportunities and
implementing innovative process improvements
because it doesn’t address the root
sources of cost. For example, consider
the process of taking meter readings and
sending bills to customers. Part of the cost
is the consequence of inaccurate billing,
which includes the cost of fielding customer
inquiry calls. Applying the accountant-
style cost breakdown above, this part
of the billing cost is buried somewhere in
call center staff and systems costs.

Conversely, when a call center agent
fields an inbound customer service call,
such as a question about moving to a different
pricing structure, the agent relies
on access to the customer information
and billing system. So in this case, part of
the cost of delivering customer service
through the call center actually lies in the
cost of providing the customer information
and billing system. In other words, the
call center systems and staff costs alone
do not reflect the complete cost of delivering
customer service.

A superior approach is to break down
cost to serve into key operational processes:
1) metering; 2) billing; 3) payment;
4) collections; and 5) customer service.

With this operational activity-based
breakdown, underlying cost causes
become more transparent and are more
readily measured. For example, it can be
determined that a portion of customer
calls relate to inaccurate bills, which
should be allocated to the billing component
of customer cost to serve. Deploying
a better, more accurate billing system
would translate into fewer calls about inaccurate
bills, thereby reducing call center
volume and reducing overall billing costs.

Is Metering a Retail Cost?

When utilities separate their retail and distribution
operations, the meters often fall
under the distribution network as it covers
field personnel that maintain the meters.
Conversely, meter-reading personnel are
often identified with the billing department
as the billing process drives the need for
timely meter reading. For calculating utility
customer cost to serve, is metering a
distribution cost or a retail cost?

Utilities in restructured markets, including
those with competitive metering, have
faced this question and, in most cases,
they have determined that metering costs
fall to the retailers, regardless of which
party owns the meters. Therefore, metering
is a component of customer cost to
serve. The Gartner Group reinforced this
conclusion when it included metering as
one of the top-level customer cost-to-serve components in its groundbreaking
benchmark study in 2006 of utility cost
to serve in the U.S.. Likewise, leading
research firm Datamonitor included
metering costs in its studies of comparative
utility customer cost to serve in Britain,
Australia and Europe.

Bad Debt – Part of Cost to Serve

Figure 1: Utility Value Chain Costs Showing
Customer Cost to ServeSeveral times a year, as a matter of policy,
a utility may write off debts unpaid for
more than 180 days. This can be considered
a financial transaction irrelevant
to customer service and, as such, could
reasonably be excluded from the operational
costs of running the customer service
department. Nevertheless, bad debt
expense is most definitely part of the cost
of doing business with retail customers
and, therefore, a component of overall
customer cost to serve.

The upside of including bad debt in
the cost-to-serve equation is collections
managers see how their receivables ledger
fits into the bigger picture. For example,
a collections department manager could
achieve savings by reducing headcount,
but it would be a false economy if the
end result is less effective collection of
overdue debt and an increase in bad-debt
write-off. Similarly, utilizing a new technology-
based service to improve collections
performance might increase collections
department costs but also
yield much greater savings by
reducing bad debt. In both of
these cases, a cost-to-serve
approach linking individual
costs to the overall business
process outcome would provide
a sound basis for decision

Percentage Breakdown of Cost to Serve
Figure 2 represents a typical
breakdown of overall cost
to serve for a large North
American utility. Next, we
will walk through each operational
area shown to see how
some true efficiencies can be

Metering Costs Likely to Increase

Metering costs include:

  • Regular field meter reading;
  • Special reads/out-of-cycle reads;
  • Meter service including meter installation
    removal, change, maintenance
    and testing;
  • Amortized cost of the meter asset; and
  • AMR/AMI and interval meter reading.

The widespread rollout of interval meter
reading has the potential to increase
costs substantially and, consequently,
to increase the customer cost to serve.
The objective will be to decrease costs
in other parts of the value chain, such as
transmission cost and wholesale costs at
peak times.

Billing Inaccuracy Compounds Costs

The American Public Power Association’s
2005 Customer Service Benchmarking
Study revealed a wide range of billing
inaccuracy rates among its utility
members, ranging from 0.004 percent
to as high as 8 percent. Inaccurate bills
drive up costs in a number of ways: more
complaints to the call center, repeat field
visits to re-read meters, back-office staff
time spent canceling and rebilling and the
additional special-run printing and postage
to send a second bill.

It makes sense to take advantage of
economies of scale in print and mail. There
are print shops that produce billions of
pieces of mail per year; at those volumes,
they can achieve efficiencies beyond even
the largest utility’s bill-printing operation.
But even more substantial cost savings
can often be found by re-examining the
overall billing process and the relationship
between billing and payment and the
consumer. Billing triggers customer calls,
so it follows that sending fewer frequent
bills (by moving to bimonthly or quarterly
billing, for example) can reduce call center
volumes. The lower costs of e-billing are
also well-documented.

A major chunk of the billing cost relates
to the number of staff required to operate
billing systems and resolve billing
exceptions. For example, a business task
such as a rate adjustment might require
several staff to spend several days using
one legacy billing system. But a billing
software package with more sophisticated
automation might only require a 10-minute
configuration and validation.

Low-Cost Payment Methods and Channels

The payment cost component of utility
customer cost to serve is an average
across all of the payment methods offered
by the utility and used by its customers.
Utilities in the U.S. have, in general, been
slow to offer the convenience of wideranging
payment options, with regulatory
restrictions often cited as the barrier to
implementing more expensive methods
such as credit card payment. Consider the
high number of utility payments still made
by check in the U.S. – up to 80 percent
(UtiliPoint, 2007) – with associated costs
of return postage and remittance processing.
Contrast that figure with other industries
or, indeed, utilities around the world
where, in some cases, 100 percent of payments
are made electronically, avoiding all
postage and remittance processing costs.

There can be wide variations in actual
payment costs across a customer base,
depending on the type of payment
(check, cash, credit card and electronic)
and the payment channel (over the counter,
call center, post, website, agency,
bank and payment service). Another
factor is payment frequency. The annual
payment processing cost for a customer
that pays four times a year using an
expensive channel can be less than the
total payment processing cost for a
customer that uses a cheaper payment
method but pays weekly.

An additional consideration is the relationship
between payment processing
and collections and bad debt costs. Some
payment methods may seem expensive
on the surface but offer substantial
benefits in reduced collection and bad
debt expense. For example, in the case
of credit card payment, the card issuer
bears most of the bad debt risk.

Combating Collections Costs

Figure 2: Cost-to-Serve Breakdown
for a Typical Large UtilityIt is estimated by information services
company Chartwell Inc. that more than
$1.7 billion in revenue is written off by utilities
in the U.S. every year, an average of
$8.50 uncollected per customer.

In addition, up to 40 percent of call
center agent time is spent on overdue
payment, payment arrangements or collections
activities. Standard utility
practice is to issue reminder letters when
debt reaches a certain number of days
past due, and these reminders
are a major trigger of calls to the
call center. Legacy billing systems
are often not designed to handle
complex modern scenarios, such as
multijurisdictional regulatory constraints
on collections activity or
tailored collection paths for different
segments of a customer base.
So collections efficiency suffers and
bad debt grows, which both result in
increases to the cost to serve.

Customer Service Costs Reflect System Usability

Absent more sophisticated costto-
serve analysis, a utility might simply
divide its customer base by its number
of call center agents to gauge whether it
has a high or low customer service costto-
serve component. It is not unusual for
some incumbent utilities running legacy
systems to have 100 call center agents
for every 100,000 customers while new
start-up entrants with better technology
and processes may have as few as 10 to
20 agents serving the same number of

This was the case for a European
retailer with 2 billion customers. The
government forced the retail department
to cut all ties with its distribution
affiliate, which revealed its high customer
cost to serve of more than EUR100/customer.
The newly separated retailer was
forced to reinvent itself from scratch
with advanced systems and business
processes in order to compete effectively
against lean, aggressive new entrants
with much lower cost to serve.

Poor customer information system
usability will also drive up customer service
labor costs. Some systems require
users to traverse 10 or more screens to
complete a common transaction, such as
signing up a new customer, while others
provide a single screen with relevant customer
information consolidated to respond
to 80 percent of calls. The best systems
place a high priority on usability engineering
and support utility-specific customer
relationship management (CRM) functions,
thereby eliminating the need for separate
billing and CRM systems.

Driving Down Cost to Serve With Meter-to-Cash Outsourcing

Certainly utilities can lower their cost
to serve through system and process
improvements in areas of metering, billing,
payment, collections and customer service,
but it is by outsourcing the full meterto-
cash transaction cycle that utilities can
radically reduce their cost to serve.

There has been tremendous attention
paid to major utilities outsourcing their
customer service operations, such as TXU
to Cap Energy and Nicor to IBM. But these
business process outsourcing (BPO) deals
tend to emphasize the financial engineering
aspects in order to yield short-term
financial benefit. Even though service-level
agreements are in place, the metrics used
typically focus on operational basics and
motivate the BPO vendor to save costs in
the near term rather than seek fundamental,
long-term improvement. They tend to
be motivated to do the same things more
cheaply, rather than to examine whether
the same tasks will actually be necessary
as business objectives, the industry and
technology evolve.

Under a new transaction-based model
of outsourcing, however, the outsourcer
owns the technology assets and intellectual
property underpinning its services.
Thus, the outsourcer can combine the
best technologies and processes in ways
that achieve far greater efficiencies and
cost reduction. This new model also provides
an outsource option for transaction
automation while allowing the utility to
retain control of its vital call center customer

If utilities are to determine their customer
cost to serve accurately, they
need to approach the equation by looking
at key operational areas, not by dividing
costs according to department. It is from
this perspective that they can then identify
new ways of achieving cost savings
and as well as increasing efficiency.
Information technology and the correct
type of outsourcing model are two fundamentals
in helping utilities realize significant
cost-to-serve savings and start
to make a dent in the $1 billion waiting
to be saved.