Note Number: DF-14-9264
Customer Information Is the Lifeblood of CRM
13 December 2001
John Radcliffe | Kimberly Collins | Jennifer Kirkby
Customer information is the basis of customer insight and effective customer interaction. It must be sourced, managed and deployed strategically.

PDF  103185.pdf
anchor arrow Topics

Customer Information Is the Lifeblood of CRM

Customer information is the basis of customer insight and effective customer interaction. It must be sourced, managed and deployed strategically.

Bottom Line

Key Issue
During the next five years, how will business strategies, behaviors, processes and technologies evolve to enable enterprises to develop more profitable customer relationships?

Enterprises need a CRM information strategy that defines how the enterprise intends to source, manage and deploy its customer information assets, and it must be owned and executed by someone with sufficient clout. Otherwise there will be "customer information anarchy," resulting in a glut of unneeded data and a famine of vital information. Most enterprises' CRM information capabilities are poor the result of numerous and fragmented departments, initiatives, databases and systems. Such fragmented approaches hurt the enterprise in many ways:

  • The costs of storing and managing duplicated data
  • The inability to handle customer interactions and contacts efficiently, effectively and accurately
  • The lack of insight into the customer's current and potential value
  • The lack of insight into the customer's past and likely future behavior and requirements
  • The lack of ability to properly segment and profile customers for differentiated product offerings and service levels
  • The inability to calculate ROI or measure CRM strategy success

Customer information is the foundation of any CRM program and a proper strategy is essential for success. This Research Note, which relates to Gartner's Eight Building Blocks of CRM framework (see "Eight Building Blocks of CRM: A Framework for Success," AV-14-9265), focuses on establishing a customer information strategy.

Creating and Maintaining Data Quality

Enterprises commonly assume that their customer data is accurate enough to support CRM initiatives, but lack of quality customer data is one of the top causes of CRM failure (see "Seven Key Reasons Why CRM Fails," COM-13-7628). Enterprises are often unaware of the scale of the problem and the cost, time and resources needed to correct it. There are a range of data quality tools (see "Data Cleansing: Detergent for CRM," TU-13-8625) that identify quality problems, re-engineer the data, standardize names and address and identify relationships (e.g., household and business partnerships). These tools are usually employed following extraction of the data prior to loading into the analytical systems, which is when most semantic and data inconsistencies first appear. However, these quality problems are actually created in the operational systems. Therefore, it is sensible to also tackle the entry points in the operational systems, in addition to maintaining vigilance at the entry points to the analytical systems.

Action Item: Data quality is an ongoing process that spans all systems that create and hold customer data. Enterprises must introduce data administration policies and processes close to the point of data capture. This includes identifying data stewards and giving them incentives to maintain that quality.

Enabling Consistent, Integrated Customer Interactions

Creating an integrated, multichannel, customer-facing view for customer interaction: Customers demand easy and convenient access via the channel(s) of their choosing and the creation of multiple touchpoints to serve the customer is central to meeting this demand. Unfortunately, multiple "touchpoints" also increase the potential for inconsistency in interactions. Enterprises that request the same customer information several times or cannot reconcile the fact that the same customer has visited different touchpoints, or is related to other members of the organization in a B2B world, will create dissatisfaction. A single customer database for operational purposes may seem the ideal solution from an architectural and data management viewpoint, but it is unlikely to be feasible for most large enterprises due to the diversity of their applications portfolio and associated databases. Strong efforts should be made to minimize the number of customer databases, and integrated, cross-functional, multichannel customer view is one of the major advantages of CRM application suites. However, in many cases, it will have to be integrated in a pragmatic fashion via application integration middleware to multiple separate application systems, each managing a portion of the overall "federated customer database," to create the perception of a multichannel, customer-facing view at the touchpoints.

Action Item: Enterprises must create an integrated, multichannel, customer-facing view, but be pragmatic about how to achieve the right perception at the touchpoint.

Integrating data from all relevant parts of the enterprise (and beyond) for customer interactions: When planning the customer information strategy, it is essential to think holistically about customer interactions. Data that is relevant to setting customer expectations, making commitments, determining progress toward meeting those commitments and delivering the commitment needs to flow between systems. Although vital, the touchpoints are only part of the story. Promises made and expectations set at the touchpoint (e.g., delivery of a product, correct and timely billing, provision of a service) are dependent on other departments (e.g., product units, marketing and finance) and their application systems. They may also depend on information from business partners and their systems. Therefore, any holistic customer information strategy has to include consideration of all systems that have an impact on the customer (e.g., ERP, SCM packages, retail banking core banking engines and telco billing systems) and there needs to be consistency of customer identification and addressing details across those systems.

Action Item: Enterprises need to think holistically and consider all systems beyond the touchpoint that have an impact on customer interactions.

Creating Customer Insight

Determine what information is needed to create the insight: Customer insight or understanding is created by analyzing customer and market data. The first steps are to determine:

  • What customer-specific insight (e.g., behavioral or value information) is needed to plan and optimize customer relationships
  • What customer strategic metrics (e.g., retention or profitability rates) are needed to measure success in meeting CRM strategy objectives

The definition of the "customer" and hence the insight and metrics required will differ greatly between B2B and B2C. In B2B, "customer" could apply to the enterprise and its employees. In B2C, the "customer" is usually the individual, but could also be the household. The second step is to take an inventory of available data sources, determining where the necessary underlying data can be collected internally (and how to handle duplicate inconsistent potential sources) or where it can be obtained externally. By working backward from the business goal, it is possible to concentrate on sourcing the right data, as opposed to collecting and analyzing irrelevant data simply because it is readily available. It will probably also be necessary to impute certain pieces of data; consideration needs to be given as to where to source the necessary base data.

Action Item: Enterprises need to be driven by their customer insight business needs, not by the availability of data.

Obtaining and transforming the data to build the customer view: To create the analytical customer view, data must be extracted from the disparate data sources, transformed into a consistent (as different systems may have different syntax and semantic rules) and usable (as it may be codified) format and integrated with other data ready for loading into the target analytical data structures. ETL tools can be used for these tasks. They are data-set-oriented, tend to be batch-based for bulk data transfer and have strong metadata support for transformations. If there are more real-time requirements, an integration-broker infrastructure may be used to pass key event data from the operational to the analytic environments. In addition to collecting operational data internally, it is also likely that external data, from vendors such as Acxiom, Experian and Dun & Bradstreet, can complete the gaps in coverage and add to the overall customer view. This is particularly true in a B2B environment. In a combination of technology, software, processes and services called customer data integration (see "Customer Data Quality and Integration: The Foundation of Successful CRM," R-14-7181) a more coherent picture of the customer relationship can be built up. This is made up of the three basic information types (see Figure 1):

  • Descriptive data (i.e., demographic, organizational and industry classification data)
  • Behavioral data (i.e., details on products held and interactions)
  • Contextual data (i.e., relationship drivers such as satisfaction, competitiveness and attitudes)

Figure 1

Customer Information Types


Source: Gartner Research

Loyalty schemes are also an effective method for collecting and maintaining data on customers (see "Customer Loyalty Programs: The Next Generation," COM-13-8246).

Action Item: Enterprises need to plan how to build up the required coherent customer view from both internal and external sources, using appropriate tools to obtain and transform the data.

Storing customer data for analytic purposes: The foundation of any serious enterprisewide strategy for acquiring, analyzing and distributing customer information is a data warehouse (see "A Taxonomy of Database Implementation Styles," TU-13-5058). Although the term is often used loosely, a true data warehouse has a cross-functional structure, with an application-neutral data model. It can therefore provide a long-term basis for a wide range of requirements. In contrast, data marts are designed to address a unique set of data analysis requirements. They are likely to perform well in meeting those design requirements, but have difficulty meeting evolving demands that require different data structures. Data marts may be dependent on the data warehouse routinely feeding it with data from it or may operate independently, acquiring data independently and running the risk of creating an inconsistent picture.

In an ideal CRM scenario the central data warehouse would be the focus of any analysis activity, but it will often have to be augmented by data marts (see "The Role of the Data Warehouse in Touchpoint Support," TU-09-7804). This may be because the particular analytic tools (e.g., a data mining tool) require specialist filing systems or it may be because the marketing applications of the chosen CRM vendor (e.g., E.piphany and Siebel) require that the customer data be in a data mart with a defined schema.

Action Item: Enterprises should view the creation of a true data warehouse as one of the key foundations of their CRM information strategy, but should recognize the need to introduce complementary data marts for specific purposes.

Analyzing customer data and creating insight: CRM analytics are currently in vogue and are at the peak of inflated expectations of the Gartner 2001 CRM Hype Cycle (see "The 2001 CRM Hype Cycle," T-13-0753). Analytics vendors of all types are attempting to leverage the "CRM bandwagon," as they realize that effective interaction is dependent on insight. To aid the selection of the appropriate technology, Gartner has defined three types of CRM analytics tools (see "The Three Categories of CRM Analytics," T-14-5545):

1. Historic analysis is useful for analyzing the success of business operations or past customer behavior. It manipulates data to see trends and patterns, but relies on the analyst to decipher the meaning.

2. Customer centric analysis is predictive, creating a deeper understanding of the customer's potential relationship with the enterprise in terms of propensity to buy or churn and lifetime value.

3. Market analysis draws out non-obvious themes and affinities, such as customer segments and buying patterns.

CRM analytics enable enterprises to plan relationships at the strategic level and optimize them at the individual level.

Action Item: Enterprises should understand the capabilities of the different CRM analytics tools and select and use them appropriately (see "Benefits and Pitfalls of Embedding Analytics Into CRM," AV-14-7182).

Applying Customer Insight to Customer Interactions

Customer relationship optimization: Enterprises are increasingly using customer insights to optimize customer relationships (see "Customer Relationship Optimization What's in a Name?" M-13-5997). These interactions can take three forms:

1. Preplanned outbound communications: B2C campaigns are now potentially multichannel, but direct mail still dominates. Outbound communications are still largely product-centric (i.e., the enterprise wanting to sell a product or service) as opposed to customer-centric (i.e., the customer needing something, based on his current situation). As such, they still suffer from low response rates due to lack of relevancy to the customer at that time. Historical and predictive analysis can be used to improve relevance and receptiveness.

2. Event-driven outbound communications: This type of interaction aims to be more relevant to the customer because it relates to a lifetime event (e.g., moving house), a series of events or nonevents (e.g., a failure to use the service) or an external event (e.g., bad weather affecting travel plans) that have affected the customer. Analysis is needed to understand the customer situation, deduce the likely relevant offers and predict the likely behavior. In addition, there is the need to actually "trap" the event and trigger the outbound interaction within the time window of opportunity. This is the greatest challenge.

3. Guiding incoming interactions in real time: Some Type A B2C enterprises (see Note 1) are applying predictive models to the customer's details at the start of, or during, an interaction. This allows them, during the interaction, to determine the appropriate retention proposition or the best cross-selling offer, based on real-time analysis of the propensity of the customer to churn or the propensity to buy.

Note 1
Type A, B and C Enterprises
Type A enterprises are technology-driven, and are often willing to risk using immature, cutting-edge technologies to gain a competitive edge.
Type B enterprises are moderate technology adopters, using new technologies once they have been proven and have entered the mainstream.
Type C enterprises, which are technologically risk-averse and cost-conscious, are usually among the last to adopt new technologies.

All the above interactions are potentially multichannel and any insight developed through CRM analytics must be able to be deployed to any channel. The underlying goals are to use the derived insight to make the individual communications and interactions more relevant to the customer as part of an ongoing set of conversations where the relationship is strengthened and value is delivered and received.

Action Item: Depending on the business model, enterprises should think in terms of optimizing customer relationships across the three styles of interaction, determining which initiatives are best supported by each. This will require planning what customer insight is needed at what points in the process and in what locations, plus the necessary processes to produce and deliver that insight in a timely fashion.

Bottom Line

Successful CRM demands the creation of a customer information "blood supply" that flows around the organization, and tight integration between operational and analytical systems. Customer information and insight must reach the touchpoints to support consistent customer interactions, prioritize customer interactions and drive more profitable customer relationships. Enterprises that establish a business plan for sourcing, maintaining and leveraging their customer information assets and establish ownership of the issue are more likely to achieve their CRM goals and objectives and gain a competitive advantage.

Acronym Key
B2B     Business-to-business
B2C     Business-to-consumer
CRM     Customer relationship management
ERP     Enterprise resource planning
ETL     Extraction, transformation and loading
ROI     Return on investment
SCM     Supply chain management

This research is part of a set of related research pieces. See AV-14-9265 for an overview.

Customer Relationship Management; Customer Data Utilization; Customer Information Management and Application
Document Type (Best practices & Case studies)
Entire contents 2001 Gartner, Inc. All rights reserved. Reproduction of this publication in any form without prior written permission is forbidden. The information contained herein has been obtained from sources believed to be reliable. Gartner disclaims all warranties as to the accuracy, completeness or adequacy of such information. Gartner shall have no liability for errors, omissions or inadequacies in the information contained herein or for interpretations thereof. The opinions expressed herein are subject to change without notice.
Resource Id: 350467