Do the Right Thing: Gain the Customer Experience Advantage Over CRMDo the Right Thing: Gain the Customer Experience Advantage Over CRM

The term Customer Relationship Management (CRM) conjures up the idea of plotting strategies to get closer to customers while manipulating those relationships to the company’s benefit. Today's reality is somewhat different, as customers show little loyalty to any one company as they turn to the Web to find the deals that suit them best.

The ability to "manage" customer relationships turned out to be elusive at best, and expectations for CRM products were not fulfilled because they focused mainly on making internal sales processes more efficient: hardly "customer focused." CRM systems failed to provide the promised 360-degree view of the customer primarily because they didn’t have access to enough data.

Taking a more holistic approach, Customer Experience Management (CEM) sets out to ensure that each customer interaction is as positive as possible, so customers feel more inclined to remain loyal. CEM is a combination of revised processes and a more integrated technology architecture that manages customer interactions consistently across all business units and all communication channels. CEM also seeks to put every interaction into the context of all previous interactions with the customer and within a defined strategy for each customer segment so, for example, long-standing, high-value customers would typically be treated differently than transient, lower-value customers.

CEM is founded on the premise that without a single, “golden” source of data and a 360-degree view of the customer, it's impossible to make each interaction with a customer as relevant, positive and mutually beneficial as possible. Once a single view is established for all channels of communication, a customer reaching a call center will receive the same account information they might have just viewed on the Web.

This article discusses the challenges of transitioning from a CRM approach to a CEM approach. It involves spanning business units and creating and sharing a single source of customer data, which is never an easy task, but it's not an overly ambitious objective. But this article puts forward pragmatic steps and advice toward delivering a better and more consistent customer experience.

CRM and the Elusive Single View

The 1980s saw the beginnings of CRM and the suggestion that companies should become more customer-focused. The vision was that companies should break down their internal business-unit silos and focus instead on the customer. Not surprisingly, technology vendors soon offered CRM applications that encapsulated these ideals and proposed to manage the associated transactions and data. But breaking down silos turned out to be harder than most companies expected, and people soon forgot that CRM was supposed to span all the business units within the company.

The profusion of separate business units has led to disjointed processes and disparate computer systems, each with its own data sources. Finance has its enterprise resource planning (ERP) systems with customer invoices and payments, marketing has its prospect and customer databases, sales has its target accounts — the list goes on. To make matters worse, a lot of customer data is in unstructured forms, such as letters, e-mail messages, Web transcripts and recorded telephone calls. With all these sources, each managed by different systems, it's not surprising that companies lost track of the “big picture” (see "Customer Needs By Channel" chart at right).

One of the first solutions envisioned was the data warehouse, a central repository that was supposed to help solve the silo problem, but many companies soon realized that with customer data in so many sources, it was impossible to guarantee that the warehouse contained all the data needed to produce accurate, consolidated reports.

The original selling point for CRM products was that they would provide a single, 360-degree view of the customer; a report that would encapsulate all information about customers spanning all business units and activities. Sadly, CRM couldn't deliver on this promise because the original products had limited access to data — in most cases an internal data store with only a partial view of the customer.

The Basics of CEM

Companies interact with customers through many different business units and channels of communication. Marketing targets existing customers and customer prospects with different types of campaigns. Field service engineers meet with customers to deliver and service products. The contact center takes most of the follow-up queries and complaints. Through it all, telephone calls, e-mail messages, faxes, Web-based interactions, text messages, instant messages, and even pictures and videos are exchanged. The totality of all these interactions creates the customer experience — the customers’ perception of the company.

Rather than trying to manage the relationship, CEM seeks to ensure that each interaction is handled within the context of a strategy for that customer or, at the very least, a strategy for the segment that customer represents. CEM also seeks to ensure that each interaction, regardless of channel, is handled in a consistent manner that is based on the same information. It thus requires consistent customer processes that are applied enterprisewide. In the IT domain, CEM demands a more integrated (typically services-oriented) architecture that links existing and new systems into a single source of customer data.

CEM Process Framework. Delivering a consistent response in every customer interaction requires processes that span business units. The idea is that each business unit (including the contact center) has it own processes and customer touch points. Wherever customer touch points exist, the processes need to be reengineered to be in line with the strategy for that customer or customer segment. For example, if a customer calls the contact center with a sales inquiry while his or her firm is in dispute over an unpaid invoice, the process should contain a decision point that prescribes strategy-specific actions that would differ for a long-standing customer with an account that is otherwise up-to-date versus, say, a more transient customer.

Vendors such as Consona, Neocase, Noetica, and Portrait Software offer products that support a process- and strategy-driven approach. Those implementing the software define how customer interactions should be handled before the project gets to the application development or configuration stage. Using the example above, the sales-call process would include a decision point based on the value of the customer and the state of their account before prompting the agent to go ahead with a sale.

CEM Technology Framework. When most people think of CRM, they think about a single application: Oracle (+ Siebel), Onyx (now part of Consona), SAP, or, in the SaaS world, RightNow or salesforce.com, or, in the open source world, Queplix or SugarCRM. By its very nature, CEM goes beyond a single application. It requires a technology architecture that supports all processes that touch customers, all technologies that support channels of communication, all applications and tools (including CRM) that manage customer transactions and structured and unstructured data, the data warehouse and all the BI and analytics that let companies produce a single view of the customer (see "Customer Experience Obstacles" chart below). This might seem a daunting task, but there are a few basic components that will get the journey started.

The first step is to address systems that support enterprisewide processes and systems that foster cooperation among business units; this list would include process modeling software as well as collaboration tools such as e-mail, workflow and collaboration tools.

The next step is to address the issue of customer data. Applications that manage customer data range from ERP, CRM, sales and operations planning, customer software, specialized customer care and billing products to call switching equipment, Web servers, call recording equipment and document scanners. For most companies, it's not viable to rip and replace these with a single application, so the answer lies in master data management (MDM): the creation of a master source of customer data that can be used across the enterprise.

MDM can be implemented at one of two levels: analytic MDM or operational MDM. Both begin with deploying tools that can extract customer data from different sources, load it into a single repository, reconcile inconsistencies and clean the data, and finally enrich the data using third-party sources such as postal lists or credit agency databases. Companies can remain at the analytic-MDM level, using the data source to produce customer-focused reports and analytics, or they can move to an operational-MDM level, using the source to synchronize customer data back to all the original data sources.

With so many sources and types of customer data, companies should not underestimate the challenge of initially extracting the data, a task that might require as many adapters as there are systems. Vendors such as Informatica, IBM, iWay and others offer adaptors for a myriad of data formats and applications. BI vendors such as Business Objects and Cognos also have data integration tools and adaptors, although most are designed to feed their own repositories. Specialized contact-center software vendors such as Latigent, Symmetrics, and VPI, have built-in adaptors for most commonly used systems while vendors such as AIM Technology, HardMetrics, and Merced Systems go one step further by including adaptors for CRM systems and other apps as well. Despite the wide array of pre-built adaptors available, it might still be necessary to build adaptors for home-grown systems and specialized software.

In the area of data cleansing, vendors such as Acuate, Dataflux (a SAS company), Informatica, Initiate, Siperian and Trillium offer software that supports the creation of consolidated metadata, the definition of the ultimate customer record, and well as deduplication and enrichment of data. Most of these vendors offer built-in reporting and analytics, and most will also let companies access the customer repository and use preferred tools to develop specific reports and analytics.

For companies that want to move up to the operational MDM level, Initiate, Siperian and Trillium, among others, support the synchronization of customer data back to enterprise systems.

Customer-Focused Analytics

The technology framework discussed above will deliver a single source of truth for customer data and, hence, the essential 360-degree customer view, but what are the core characteristics of customer-focused reports and analytics? First and foremost, they're about customers and business, not simply efficiency reports on how many calls were handled, the average length of call, how many calls were transferred between agents and the like. Reports should include effective measures such as which customers transacted what levels of business, average wallet share, the number of interactions resolved in the first interaction, the cost of providing customer service to different segments and other, deeper insights. Where possible, predictive analytics should be applied; for example, which types or frequencies of inquiries or complaints are likely to lead to a customer defection?

The analysis also needs to be in a format that is appropriate to the user. A call center agent, for instance, might need a simple alert to take special action, such as transferring a call to a supervisor because the account is X many dollars in arrears. A top executive, on the other hand, might want a high-level report showing how much repeat business has been generated through each channel. Most vendors identified in the "Technology Framework" section above can produce these types of reports.

Companies obviously want to retain existing customers while also selling them additional products and services. To do so more effectively, companies should close the loop on customer-facing processes by analyzing data on past outcomes and optimizing the process. Chordiant decision management tools let companies pull data from many sources of customer interaction, analyze the outcomes and use a rules-based engine to model how those interactions could be improved. The results can used to personalize the next interaction with the customer. Such techniques have been shown to create additional business and increase customer satisfaction.

The Unstructured Content Challenge

The biggest challenge in creating a single view of the customer is managing unstructured data. Until recently, data from Web interactions, interactive voice recordings and call recordings has been almost impossible to use, but this is changing. In one example, SAS has partnered with speed-trap in the UK, which has developed a tool whereby a bit of code is embedded in a company's Web site to capture every key stroke made during the interaction, be it to enter data, follow a hot link, navigate to another part of the site or even click in an empty space. [Editor's note: San Francisco-based TeaLeaf offers similar capabilities.] This data is fed into the SAS tools where it is combined with other data, such as that from CRM and ERP systems, to paint a picture of which customers have been using the site and their interests and behavior. This insight can be used not only to suggest Web-site improvements that might lead to increased conversion rates but also to trigger possible actions, such as an outbound call, to increase the number of sales closed.

Phone calls to a company, most notably to the contact center, typically represent more than ten times the volume of interactions than through any other form of communication. New compression techniques now make it possible and affordable to record all calls. Working with partners such as Cisco, vendors such as eLoyalty, NICE Systems, and Verint (the latter with its recent acquisition of Witness Systems) have developed applications that turn these unstructured recordings into partially structured records. By analyzing the tone, pitch, and content of these call records, companies can gain insight into customer behavior patterns, agent reaction to that behavior, customer satisfaction levels and even predictions as to how certain customers might behave in the future (as in, defect or buy new products).

Multi-Channel and Virtualization Challenges

The ultimate challenge for CEM is to produce an analysis of a customer’s entire history and current interactions, encompassing all communication channels. Clickfox has developed tools that take data from many different communication sources and attribute them to particular customers. This offers insight into customer behavior that spans different channels and can point to improvements in customer processes, Web sites and IVR systems. Customers have reported gains in self-service rates as well as reduced call volumes customer churn rates.

Companies are increasingly routing calls outside of the contact center (to specialists and experts) in an effort to improve first-call resolution rates and customer satisfaction levels. This is done by integrating a single source of customer data to call routing software (see "Insurance.com Optimizes the Customer Experience, One Segment at a Time"). Vendors such as Aspect, Avaya, Cisco and Genesys have products that can route customer interactions based on customer profiles, preferences and behavioral patterns, not just available agents. Such a step might also require a more process-driven desktop (such as those from Chordiant and Jacada) that can guide less-experience users through the required customer service processes.

The Journey Toward CEM

The CEM journey should start by developing a broader, deeper view of your customers. Achieving a single view might seem like a pipe-dream, but business- and customer-focused BI products (such as those from AIM Technology, HardMetrics, Merced Systems and others) are making it possible to build a more comprehensive picture that includes contact center data, application-specific (CRM and ERP) data, as well as sales and marketing data. This first step can serve as the basis for several short-term initiatives:

1. Provide contact center agents with more customer information — customer segments, life-time value, account summary and so on — so they can make better decisions on how to handle interactions.

2. Create more granular customer segments that go beyond customer value to spot high-value groups that have a propensity to spend, groups that have purchased more than one product, customer that are difficult to deal with, and so on. Specific CEM strategies should then be devised for each of these groups. For example, low-value customers might be encouraged to move to self-service channels, thereby reducing support costs.

3. Review all processes that touch customers — running sales campaigns, handling calls, replying to letters, chasing invoices and so on. Examine these processes in light of the segments and CEM strategies you've developed. For example, when long-standing, high-value customers are thought to be at risk of defecting, the process should route those customers to groups specializing in customer retention.

4. Update key performance metrics and any associated reward systems. One of the easiest ways to drive change is to link reward systems to these new, more customer-focused metrics. This will change the way people behave and lead to the desired business outcomes.

These steps four alone should improve the customer experience and raise satisfaction levels, but next steps on the CEM journey might take these parallel paths:

Enhance the picture of customers and their behavior by adding sources of data such as Web scripts, IVR scripts, recorded calls, scanned letters and so on. Next, extend the analysis of how customers interact across different channels of communication. As the picture becomes richer, each of the four steps above can be repeated to achieve even higher levels of customer satisfaction and business.

Address the quality of the customer data. The first step should be to create a cross-functional/departmental data governance board (or center of excellence) that includes all interested parties (See "The Next Wave in Data Management." This group's first step should be to identify the cost to the business of errors and inconsistencies in customer data (such as defections and lost sales). This should drive a data governance strategy, including, as appropriate, an MDM initiative that improves customer data quality and data access.

Consider virtualizing the call center by routing interactions based on customer profiles, not just agent availability or load balancing. This will increase the likelihood of the customer being handled through the channel of their choice, in a manner personalized to their circumstances, and in a way most likely to enhance their experience to the maximum extent.

Conclusion

CRM might have been a dream that was not achievable as it was first conceived. The promised 360-degree view of customers has been a long time in coming because of difficulties of accessing all the required data. CEM is a more pragmatic approach that focuses on maximizing the customer experience by ensuring that each customer interaction is handled in a consistent manner across all business units and communication channels, and within the context of a predefined strategy for each customer segment. CEM requires process change, the creation of a single source of truth about customers and a more integrated technical architecture. Pioneering companies have begun the CEM journey, but as more technology comes to market and best practices are established to support such initiatives, you can expect the adoption rate to accelerate over the next three to five years.

Richard Snow is VP & Research Director, Contact Center Performance Management at Ventana Research, a practice dedicated to helping organizations improve the efficiency and effectiveness of multichannel contact centers. Write him at [email protected].