More than Data Warehousing: An Integrated View of the Customer
That vibration on your belt is familiar, but you're still startled when the pager interrupts your session with The Wall Street Journal as you wait for a flight in the Miami airport. You glance at the message, surprised that it's from your online broker. You almost delete it, assuming it's a pedestrian banner ad from the website that has somehow wormed its way into your pager address. Then you notice that it's a personalized message. You can have the fee waived for your next trade. No way was this message sent to everyone in their database - it'd cost them a fortune! Apparently, your online broker has noticed your conspicuous lack of trading activity since you opened the account eight months ago. The offer makes you want to act immediately. In retrospect, you decide it's not the eight bucks that is spurring you on, but the poignant reminder that your money has been earning zero-point-nothing percent interest for nearly a year in the brokerage account. You make a mental note to log on when you hit the ground in Minneapolis and take advantage of the offer.
This campaign for an online brokerage firm garnered a 30 to 40% activation rate, and 20 to 30% of those people repeated. Its value - and success - in terms of increased accuracy and reduced costs are attributed to having what Accenture calls an "Integrated View of Customer" (IVoC).
Integrated View of Customer is an enabling business capability that supplies the missing link for CRM implementations: the connection between customer interaction channels and customer strategy. IVoC brings together demographic, behavioral, and contact data for analysis and customer scoring. It also makes those customers and their scores available to, for example, the brokerage's website and email system for outbound campaign execution.
In addition to gaining higher value through some very profitable campaigns, the brokerage also improved speed and increased throughput. Models for predicting response that used to take six weeks to develop were done in one-and-a-half weeks because the brokerage had an integrated view of its customers.
Gaining Greater Value
Many companies have embarked on CRM projects to improve their marketing and service capabilities - for example, to increase cross-selling and up-selling; improve call center routing, based on the type of customer calling in; and to more effectively target the customers most likely to be interested in what they are selling.
These are worthwhile objectives, to be sure, but the central problem remains: neither operational CRM databases nor data warehouses meet the needs of a holistic marketing, sales or customer service agenda. Most CRM database implementations are channel specific; most data warehouses are not channel centric, but they are not customer-centric, either. No CRM database integrates everything you need to know about your customers - their interactions, the products and services they've purchased and the channels they've used - at the exact moment you need to know it. The results? Unused data and under-utilized processing capability. The missing link is an Integrated View of Customer that is the touch point for all interactions a company has with its customers. This capability is essential for companies to realize better value from their existing CRM investments.
Equally important, as companies feel their way through initiatives meant to extend CRM value across channels and product lines, they are sub-optimizing the integration of these solutions because they lack a key ingredient - context. Defined as the intent or purpose of the customer interaction, context determines how all of these interactions work together to shape a customer's overall experience with a company. Today's cross-channel integration efforts are usually not holistic or strategic. Efforts do not consider the breadth and depth of the overall customer experience when the scope is being defined. Instead of thinking big, starting small and growing fast, managers are thinking small, starting slowly and not growing because they lack a road map.
A typical result is cross-channel integration that occurs one data element at a time. Call centers gain access to email addresses captured on the Web, and the website gains access to customer service contact data stored in the call center system. None of this takes into account the enterprise solution to such problems, nor does it address the enterprise value. Context is required to translate customer strategy into action.
A Unified Solution
An Integrated View of Customer capability that will provide this context is composed of:
o Customer contact and marketing processes that are linked to ensure that customer insight can be generated and leveraged by the customer contact channels.
o Technologies that deliver the required data to the right business function on a continual, real-time basis.
o An organization and governance model, enabled by technology, that can apply the customer strategy across business units and contact channels.
To achieve an Integrated View of Customers, a company's customer contact channels should be connected together to ensure that contact history, transaction history, demographic data and preferences are available as needed across all channels. In addition, key information required to generate customer insight needs to be captured and made available to customer insight functions, such as analytics and campaign management, in a timely manner. Finally, messages and offers intended to shape customer interactions need to be delivered back to contact channels such as the sales force, call centers and the Web.
An Integrated View of Customer capability is essential to ensure that the right data is delivered to the right business functions at the right time to help shape subsequent customer interactions. Not all customer data need to be accessible to all processes at all times - only meaningful data, delivered just in time.
This is a far cry from the data warehousing approaches of old. In fact, we see the Integrated View of Customer capability as the third stage in the evolution of data warehousing. The first stage, which came into being in the 1980s, highlighted virtual information access. It provided for online access to business information contained in systems of record. However, this approach was characterized by poor performance and inadequate data availability. Systems typically were locked down and inaccessible except for nighttime batch reporting. Online query and reporting software was limited. No architecture was available to support the concept, and no methodology or process existed for building such a system. Finally, no solution could be found to the problem of inconsistent data across multiple systems.
In an attempt to correct these deficiencies, a data warehousing methodology emerged in the 1990s. Like any evolutionary process, this step solved previous problems but created new side effects. Data warehousing methodology insisted that companies define specific business problems to solve or questions that needed answering on an ongoing basis. This approach helped guarantee business value, performance and availability for known questions. However, due to the time-intensive and capital-intensive nature of this approach, flexibility was limited. Data were delivered as snapshots - one point in time. As businesses took new directions and users changed their minds about what types of questions they wished to ask, data warehouses were slow to respond and required ever-larger development efforts.
The ultimate example of business users changing their minds occurs when they become customer centric. This represents a huge process shift for businesses and often invalidates existing data warehouses that were built to be product centric. Moreover, this process change brings with it two new requirements that traditional data warehouses are ill equipped to satisfy: value and speed. Data warehouses must be constructed more quickly, with less investment, and be flexible as businesses' customer-centric needs change. Analytical results must be fed back into interaction systems so they can deliver value by shaping the customer experience. A real-time feedback loop was never a requirement before, but it is a necessity now.
Three Critical Components
Enter the Integrated View of Customer, the third stage in the data warehousing evolution. The IVoC capability encompasses three vital components that produce the value and speed today's customer-centric business strategies require:
o Data includes the information required about the customer to extend CRM solutions across contact channels or product lines within your enterprise.
o Rules are the instructions for using the data to shape the next interaction with your customer. Rules define what you do with the information after it is warehoused.
o Context is the translation of your customer strategy into rules meant to deliver that strategy.
Data
A wide variety of data is required to drive CRM value and provide an Integrated View of Customer. More is not necessarily better, however. What you really want your IVoC capability to enable is delivery of the right data to ensure the desired customer experience across channels. For instance, if preliminary analysis shows that an e-mail address plays no part in customer segmentation or is not used in any channel other than the Web, then it is unlikely that you will want to store that attribute in your IVoC data architecture.
You also should consider how the data could best be organized. Information required to drive customer interactions, such as demographic updates, will require an interactive data architecture. The data required for generating customer insight also will need to be accessed by data mining and analytic tools.
Timing is another important consideration. For example, some businesses would find significant value in receiving real-time updates as customers log complaints. Complaint data would help drive real-time analytics and effect customer treatments applied in real time at the point of interaction. Another example: Information shared between two customer contact channels is very often required in near real time, as a customer may use both a call center or a Web site within minutes of each other.
In addition, the quality of data and privacy issues must be considered. An IVoC data architecture also must be augmented and cleansed as appropriate and, at the same time, provide an organization with a centralized record of which data can be used for which functions.
Rules
The rules required for an Integrated View of Customer capability come in two primary forms - personalization and data transformation. Importantly, their development should be considered a business function and not simply delegated to your IT department. Why? Most rules are manifestation of business objectives, decisions and requirements rather than technical implications of those requirements.
Personalization rules shape each customer interaction. They provide for consistent, timely and relevant individualized interactions and offers across multiple customer touch points. Personalization rules govern what information is "pushed" to customers and how customer information is collected. Data transformation rules, on the other hand, govern non-customer-facing activity. They provide for a similarly important business function: merging customer data from source systems or third parties. Both forms of rules - personalization and transformation - are most effective when they address the broadest possible set of scenarios.
Context
Context is typically the missing piece for many CRM implementations, yet it has the most impact of all three IVoC components. Two facets of context are especially important:
o A holistic understanding of all the potential customer interactions within an organization, the events that make up those interactions and how the outcome of each interaction affects subsequent interactions or the overall customer relationship.
o The reason why a rule is being implemented or a decision is being made.
To better understand the critical importance of context, think back to our brokerage example at the beginning of this article. Automated stock alerts sent by a brokerage to PDAs, cell phones or pagers may have the unintended effect of driving trading requests to a call center if convenient Web connections are not available at the time customers receive the trading message. The call center, not anticipating such peaks, may experience reduced service levels at the very moment their customers desire and expect immediate attention. After all, it was the brokerage that notified them, not the other way around. It is at this touchpoint that context becomes real and urgent.
In addition, the purpose of each rule needs to be determined. As more and more rules governing interactions and data transformation are implemented across the enterprise, it is important to understand the intended effect. Which rules are meant to contribute to which dimension of the customer strategy - acquisition, cross-selling or retention? As the strategy is defined, rules should be defined for each of the dimensions. For instance, an up-selling rule implemented as part of a direct mail campaign for an automotive manufacturer may actually cannibalize a retention program currently in place at the customer service center. The company's IVoC capability should provide awareness of this conundrum and the ultimate long-term strategy. This goal is the manifestation of context, which governs which of these rules should take precedence at any given time.
Looking Ahead
As you consider better ways to link your customer interaction and strategy, the following questions will help you decide whether you need to develop an Integrated View of Customer capability, or, if you already have an IVoC capability, whether it requires improvement.
Customer Relationship Management
o Can I share customer contacts across channels?
o Can I share a customer's business transactions across channels?
o Can Customer Service Representatives (CSRs) view a customer's entire book of business?
o Can customers view their entire book of business on the Web?
o Can Marketing analyze a customer's entire book of business?
o Can Marketing analyze complete and consistent attitudinal and behavioral data for a customer?
o Can I run an integrated campaign?
o Can I run a multi-channel campaign in real time?
o Can I differentiate the level of service provided based on who is calling?
o Can I consistently identify a customer on any channel?
o How many callers can a CSR accurately identify?
o How many times does a CSR have to enter in the customer number?
o Do my CSRs know my customers' Web preferences?
o How many of my customers use two or more channels to interact with us?
o How much revenue is generated from a cross-sell by a CSR?
o Can I cross-sell and up-sell on the Web?
Customer Data
o Do I have the right kind of data? Is there enough information and is it rich enough? Could it move a metric, such as revenue per call?
o Where does all my customer data reside? Is information in the right places?
o What information is duplicated in multiple databases? How many databases?
o Is information easy to access for CSRs, marketers and others who need to use it?
o Of the shared information available, how much is being used? For example, are CSRs using the data available?
o When the shared information is used, how are revenues affected?
Rules
(instructions for using the data to shape the next customer interaction)
o Are the rules consistent across multiple channels?
o Can I deliver a consistent message across our channels and make sure that I'm doing so?
o How many systems store rules that govern customer interactions?
o How many systems store rules that unify customer data?
Context
(translating your customer strategy)
o Am I meeting the goals set forth in my customer strategy?
o Am I experiencing undesirable side effects from my campaigns?
Tough questions, to be sure, but they are necessary to ask in order to unlock the value of your existing CRM investments and carry those capabilities - and your customers - to a higher level of responsiveness and profitability.
