United States
   Environmental Protection
Customer Complaint Surveillance Primer
For Water Quality Surveillance and Response Systems
Office of Water (MC 140)
May 2015

                             Customer Complaint Surveillance Primer
 A Water Quality Surveillance and Response System (SRS) provides a systematic framework for
 enhancing distribution system monitoring activities to detect emerging water quality issues and respond
 before they become problems.  An SRS consists of six components grouped into two operational phases,
 surveillance and response. The surveillance components are designed to provide timely detection of
 water quality incidents in drinking water distribution systems and include: Online Water Quality
 Monitoring, Enhanced Security Monitoring, Customer Complaint Surveillance and Public Health
 Surveillance. The response components include Consequence Management and Sampling & Analysis,
 which support timely response actions that minimize the consequences of a contamination incident. The
 Water Quality Surveillance and Response System Primer provides a brief overview of the entire system
 (USEPA, 2015).
This document provides an overview of the Customer Complaint Surveillance
(CCS) component of an SRS. It presents basic information about the goals and
objectives of CCS in the context of an SRS.  This primer covers the following
four topics:
       Topic 1: What is CCS?
       Topic 2: What are the major design elements of CCS?
       Topic 3: What are common design goals and performance objectives
        for CCS?
       Topic 4: What are cost-effective approaches for CCS?

Topic 1:  What is CCS?
CCS consists of information management systems, processes and procedures that collectively compile,
track and analyze water quality-related customer complaints indicative of a water quality incident.

Figure 1 illustrates the Funnel/Filter/Focus
surveillance approach of CCS. First, all complaints
arefunneled into one location, such as a call
management system, to ensure that complaints are
not missed.  Next, water quality complaints are
filtered out from non-water quality complaints by
Customer Service Representatives (CSR) or other
water utility staff. Finally, water quality specialists
focus on the remaining complaints to assess
whether the complaints are related to a water
quality incident or to system operations, such as
main breaks or maintenance.  CCS can track the
time, number and location of complaints that are
entered into call or work management systems, and
alert utility personnel of unusually high call
volumes or spatially-clustered complaints.
                                                Figure 1. Recommended Funnel/Filter/Focus
                                                Approach for Utility-managed Customer Calls

                            Customer Complaint Surveillance Primer

Topic 2:  What are the major design elements of CCS?
The major design elements for CCS are shown in Figure 2 and described under the remainder of this
   and Analysis
Algorithms      Location     Timing
                  Mobile Device
Figure 2. CCS Design Elements
               Central Control
Other Remote
Complaint Collection
A variety of methods are available to funnel all customer calls to
one point of contact. For example, a unified call center with a
widely publicized telephone number helps to ensure that the
majority of complaints are captured.

Additionally, procedures should be put in place for water quality-
related complaints that are initially received by external agencies,
such as a city-wide call center or a 311 system. These procedures
funnel calls to the CCS centralized complaint management system,
ensuring robust surveillance by accounting for all calls.
                                                 DID You KNOW?

                                           Many utilities with CCS utilize a
                                           311 system for complaint
                                           collection. This allows them to
                                           build on an existing system,
                                           with only minor procedural
                                           changes needed to implement
                                           the design element.
Complaint collection has two design sub-elements:

       Communicating Water Quality Concerns:  Activities implemented to ensure that customers are
       aware of how to report their water quality concerns to the utility.
       Consolidating Water Quality Complaints:  Systems or procedures that filter water quality
       complaints to a central location, facilitating timely and efficient data analysis.

Information Management and Analysis
A key requirement of CCS is the ability to systematically track water quality complaints from receipt to
closure. Existing customer complaint processes used by a utility can typically be leveraged to develop a
CCS information management and analysis system.  One of the most important decisions when
implementing CCS is determining which datastreams should be incorporated. Most of the CCS tracking
mechanisms work behind the scenes,  limiting interference with day-to-day operations.  Complaints are

                              Customer Complaint Surveillance Primer
continuously analyzed for information indicative of a water quality incident in the distribution system,
such as an unusually high numbers of calls or clustering of complaint locations.  CCS information
management and analysis has six design sub-elements:
      Complaint Descriptive Data and Categories: Capture and categorize complaint descriptions
       for the purpose of data analysis and alert investigations.
      Detecting Abnormally High Complaint Volumes: Develop processes to identify complaint
       volumes attributable to a significant change in water quality.
      Timeliness of Detection: Ensure data is reviewed with sufficient regularity to identify potential
       water quality incidents as they develop.
      Establishing Thresholds: Develop alerting thresholds that are insensitive to normal variation in
       call volume, yet low enough to detect potential water quality incidents.
      Spatial Clustering Analysis:  Use spatial analysis to determine whether an unusually high
       volume of calls is clustered,  and to determine the area of the distribution system impacted by a
       possible water quality incident.
      Alert Notifications:  Develop reliable processes for informing utility personnel when alerting
       thresholds are exceeded.
CCS can take advantage of existing information management
systems used in a typical call management process by filtering
water quality-related complaints. Utilities without formal call
management systems or software may still reap the benefits of
CCS by: 1) streamlining the manner in which water quality-
related complaints are managed, and by 2) instituting frequent
checks of the number of water quality-related calls received
        DID YOU KNOW?
Many utilities with CCS modified
their existing data management
systems by establishing water
quality-related complaint categories
and then tracking the resolution of
these complaints through their
customer service process.
Water quality-related customer complaint data that is collected should be analyzed in a timely manner for
conditions indicative of a water quality incident in the distribution system.  This involves identifying
when the total number of water quality-related complaints is unusual compared to an established baseline.
The anomaly detection process can be automated using simple counting algorithms, which automatically
track the number of calls over a defined period of time. When the number of calls exceeds a pre-
determined threshold value, an alert is generated and utility personnel notified.

If spatial data is available, the frequency of complaints within hydraulically related areas, such as pressure
zones or service areas, can also be evaluated. In addition, mapping the location of complaints can
highlight clustering, which focuses investigation and response actions.
An example of a detection timeline is illustrated in Figure 3. The
delay times for detection shown in the middle row provide a sense of
how quickly CCS data is available. For this example, data is
extracted from existing call and work management data systems and
analyzed in near real time (every 15 minutes or less) using a simple
counting algorithm. Upon generation of an alert, notifications are
sent to investigators using an existing email server.
          DID YOU KNOW?
    USEPA's SRS Program has
    published a Threshold Analysis
    Tool that performs statistical
    analysis of complaint data to
    guide the development of
    threshold values.

                            Customer Complaint Surveillance Primer
   Person perceives
change(s) in drinking
water characteristics
                     calls utility
 CSR identifies
issue and creates
   work order
Event detection
 analyzes data
Alert generated;
notification sent
 to investigator



5 minutes


15 minutes






Figure 3. Example of a CCS Detection Timeline

Alert Investigation Procedures
CCS alerts need to be promptly investigated by utility personnel to determine whether the alerts can be
explained by known factors, such as distribution system work near the area of the complaint locations.
Example steps performed during CCS alert investigations are described below.
    1.  A CCS investigation begins following receipt of an alert, signifying an anomaly in one or more
       CCS datastreams.
    2.  Utility personnel use a CCS alert investigation checklist to guide them through a predetermined
       procedure to determine if the complaints are related to a water quality incident in the distribution
    3.  If it is determined the alert is not related to a water quality  incident in the distribution system, the
       investigation is closed and logged.
    4.  If a water quality incident cannot be ruled out, the investigation continues according to
       procedures in the drinking water utility's Consequence Management Plan.

Topic 3:  What are common design  goals and performance objectives
for CCS?
The design goals  and performance objectives established for CCS by the utility provide the basis for the
design of an effective component.

CCS Design Goals
Design goals are the specific benefits that utilities expect to achieve by implementing CCS.  A
fundamental design goal of an SRS is the ability to detect and respond to water quality anomalies in the
distribution system. In addition to this fundamental SRS design goal, other CCS-specific design goals
such as improving the level of customer service can be realized. Examples of common CCS design goals
are listed in Table 1.

                             Customer Complaint Surveillance Primer
Table 1.  Examples of Common CCS Design Goals
Design Goal
Detect water contamination
Monitor the impact of system
operations on customers
Increase the level of customer
Improve the response to water
quality complaints
CCS provides an early indicator of water contamination which may
impact the health of customers or utility infrastructure.
Some utility operations, such as changing sources and chlorine feed
levels, can impact the aesthetics of the drinking water. CCS can alert
the utility if these changes are noticed by customers.
CCS can alert the utility to distribution issues, such as main breaks,
through customer complaints. This can reduce utility response time
while providing the latest information to CSRs receiving customer
Developing CCS procedures can streamline and standardize a utility's
decision-making process when investigating customer water quality
CCS Performance Objectives
Performance objectives are measurable indicators of how well the SRS meets the design goals established
by the utility. Throughout design, implementation and operation of the SRS or its components, the utility
can use performance objectives to evaluate the added value of each capability, procedure or partnership.
While specific performance objectives should be developed by each utility in the context of its unique
design goals, general performance objectives for an SRS were defined in the Water Quality Surveillance
and Response System Primer (USEPA, 2015) and are further described in the context of CCS as follows.
      Incident coverage: Detect and respond to abroad spectrum of water quality incidents.  CCS is
       limited to detection of contaminants which alter the taste, odor or appearance of drinking water.
       Within this subset of contaminants, CCS can detect incidents regardless of the source.
      Spatial coverage: Achieve spatial coverage of the entire distribution system.  Theoretically,
       CCS has the ability to cover every customer in the distribution system. Spatial coverage is
       improved by educating customers about how to contact the utility.
      Timeliness of detection:  Detect water quality incidents in sufficient time for effective response.
       This performance objective is dependent upon how quickly data is available for analysis and how
       often the analysis is performed.
      Operational reliability: Minimize downtime for equipment, personnel and other support
       functions necessary for the component to meet the other performance objectives. Operational
       reliability for CCS is achieved by ensuring that information management and analysis systems
       continue to operate.
      Alert occurrence: Minimize the number of invalid alerts, which are not caused by abnormal
       water quality, while maintaining the ability of the system to detect true water quality anomalies.
       The balance between reducing the rate of invalid alerts while maintaining detection capabilities is
       primarily a function of the quality of the data monitored by the system and the data analysis
       method(s) used.
      Sustainability: Provide value to day-to-day utility operations and distribution system
       management that exceeds the cost to  deploy and operate the component.  Because CCS involves
       little to no physical equipment, it is relatively easy to sustain.

                            Customer Complaint Surveillance Primer

Topic 4:  What are cost-effective approaches for CCS?

Utilities can take the following simple steps to develop the foundation for CCS:
       Review historical customer complaint data, and estimate a threshold for the number of calls or
       work orders that could be indicative of a water quality incident.
       Evaluate the daily volume of water quality complaints relative to the threshold, and manually plot
       the locations of calls on a map with 'push-pins' to identify clusters.
       Establish procedures for investigating water quality complaint clusters and train staff on their

Next Steps
Visit the Water Quality Surveillance and Response Website at http://water.epa.gov/infrastructure
/watersecurity/lawsregs/initiative.cfm for more information about SRS practices. The Website contains
guidance and tools that will help a utility to enhance surveillance and response capabilities, as well as
case studies that share utility experiences with SRS implementation and operation.

USEPA. (2015). Water Quality Surveillance and Response System Primer, 817-B-15-002.