In an economy where superior customer service has become the distinguishing factor among competing companies, self-help is touted as one of the simplest and least costly ways to reduce help desk costs. 

Web-based self-help tools are mushrooming, promising to deliver excellent service while reducing costs and increasing efficiencies, and delivering enviable savings to those who implement them. 

Perhaps as importantly, these systems promise increased customer loyalty due to the added convenience and increased satisfaction. It is not surprising therefore that in surveys conducted by the Help Desk Institute and by supportindustry.com, almost half of the respondents have implemented some type of self-help facility, and over half of those who haven't, said that they intend to move to self-help within 12 months.

Joe Barkai
Founder and Principal
Diagnostic Strategies

• "78% of organizations believe providing customers with Web self-service capabilities could prevent up to 40% of front line support calls. The remainder believes the savings could be even higher." - Supportindustry.com Research

Why the excitement over self-service? The answer appears simple enough: if customers can solve their problems on their own and never call the help desk, the call volume will decrease, and since the cost of a Web-based service transaction is significantly lower than engaging a human agent, this reduction in call volume translates into immediate savings.

The reduction in call volume also means lower workload and burnout, resulting in less stressed agents that provide better service. Moreover, because agents take fewer support calls, they are now free to take more revenue-generating calls.

But to the surprise of many, these logical expectations often turn into disappointments and even utter failures. The same supportindustry.com study indicates that a mere 50% of the call centers that implemented self-service have experienced reduction in call volume. Forrester Research reports that only 36% of organizations noticed a reduction in call volume, and 17% have actually measured an unexpected increase in the number of calls.

Call volume is not the only performance metric that self-help does not always improve as promised. Research shows that other key indicators, including average speed of answer, resolution time and first contact resolution rate did not appear to improve, and in some instances have actually worsened.

What Happened?
Let's begin by examining the holy grail of self-help: call volume. Self-help proponents maintain that instead of calling the help desk, customers can find the answers to their questions on the Web, thereby "deflecting" calls away from the call center.

So why didn't the call volume decline? There is the obvious answer: customers couldn't find the answer to their question on the Web and ended up talking to a live agent after all, often irate and needing more attention following their frustrating Web experience. But if this was the only reason, then identifying the missing information and enhancing the content of the self-help facility should be an easy task. However, careful examination of the detailed operation of a call center reveals additional challenges in reducing call volume by "deflecting" calls.

Capacity
Most call centers are not staffed to answer 100 percent of customer requests. Instead, they are somewhat understaffed and plan for a certain percentage of abandoned calls. In other words, in a typical call center, the demand for service always exceeds capacity, so as the self-help facility eliminates some calls, the available capacity is filled immediately by this demand.

Call Distribution
A subtle, yet profound reason for the unexpectedly small effect self-help has on call volume stems from the impact of self-help on call distribution. In a typical call center environment, agents handle a mixture of call types: some are very simple and are answered almost instantaneously because the agent can convey the information to the caller succinctly and quickly. Other calls are more complex and take longer to resolve because they require research, may need more than one conversation with the customer, and sometimes are escalated to a higher support level. The distribution of call handling times is not uniform: the number of simple and short calls is typically much greater than the number of complex, time-consuming calls. In fact, most of the calls are actually much shorter than the average call length, and there are relatively few calls that take significantly longer than the average call.

Web-based self-help facilities focus on the frequent simpler and shorter calls, which customers can easily research and resolve, but their impact on the very long calls is minimal. Once the majority of the short calls are eliminated, the overall call volume may decline, but the relative portion of complex calls increases, and things start to go wrong.

Many call centers are structured in tiers, with Tier One clearing the larger volume of the easier and shorter calls, and escalating the more complex calls to Tier Two. With the new call distribution, there are fewer calls that Tier One can solve and the percentage of calls that require escalation increases. Service level reports will show degradation in first contact resolution rates, increase in the percentage of escalated calls, and lower utilization because of lower call volume handled by Tier One. Overall, the performance of Tier One following the implementation of self-help will appear worse, and the additional capacity created by "deflecting" calls will be greatly underutilized.

Can Self-Help Help?
This is not to say that self-help can never help. Self-help can provide valuable help to customers who need the kind of information that can be delivered effectively using self-help technologies, such as providing order fulfillment status or answering frequently asked questions. One notable example is Pitney Bowes, where a self-service system to answer questions concerning postage rates increase reduced the volume of calls to live agents by 50% relative to the previous rate increase.

• "By 2005, more than 70 percent of customer service interaction for information and remote transactions will be automated." - Gartner Group

Providing technical assistance such as problem troubleshooting using self-help tools is more challenging. Simple troubleshooting information for simple devices, such as small printers, and frequently asked technical questions are reasonable applications, but when the task requires extensive technical knowledge and accuracy, and the risk of a user error is too high, self-help may not be the recommended approach.

The Opportunity
Whether a company's objective is to reduce the load in a busy call center or to streamline sales and order tracking interactions with customers, the goal of self-service technology remains the same: to leverage available knowledge resources and improve quality of the service, while reducing costs.

Almost every discussion about the benefits of self-service lures the reader to rosy payback predictions based on the high cost of a call handled by a live agent versus the very low cost of providing the same service via the Web. However, an ongoing study of call costs by Diagnostic Strategies demonstrates that this data is highly ambiguous and cannot be used as a reliable method to estimate savings.

Self-help cannot focus on cost-cutting alone. We must remember that the more the customer has to do, the less the service organization does, an attitude that may quickly and unnoticeably lead to poor service. Self-service and good service can become an oxymoron, especially when the primary focus is on cost reduction.

Users perceive the value of the self-help facility based on several, often conflicting aspects, which include ease of access, the quality and the relevancy of the information it provides, and the timeliness of updates. The following steps will help you devise an approach to implementing a self-help facility.

Determine the Need and the Opportunity:There must be a clear and recognized need that self-help can address. Not all customer issues are best addresses by self-help, nor are all user communities good candidates for this approach. The need, the opportunity, and the willingness and ability of users to solve their own issues should be defined and quantified, and weighed against the risks before any significant investment in the project is made.

Define the Value: Oddly enough, many organizations are strong believers in the "if you build it they will come" philosophy. Self-service must have a clear value not only to the implementing organization but also to the intended audience. The value can have different facets, tangible and intangible, tactical and strategic, but without establishing the customer value and maintaining it throughout the life of the service, the chances of creating a long-term success are greatly diminished.

Popular Practices vs. Best Practices (or One Size Doesn't Fit All): The fact that other organizations have supposedly reaped substantial benefits from self-service doesn't make their experience immediately apply to other situations or organizations. Many of the hyped-up success stories serve the interests of software vendors and analysts more than of the customers. Truly successful organizations have invested significantly to understand the value of self-service and how to best deliver it to their customers and sustain the value over time.

More isn't Better: A view taken by many organizations is that more information is better, essentially drowning customers with large amounts of poorly organized and often irrelevant information. Information provided to users must be easy to navigate and find, and the answers relevant and easy to understand. Customers who spent time sifting through useless information are much less forgiving than those who quickly realize that the information they are looking for is not available and are given an easy way to contact the call center.

Select Appropriate Technology: Many organizations fail because they first select a technology and then base their self-help strategy on it. There are numerous technologies that can be harnessed to provide customers self-help, ranging from Web portals access to well-organized quality documentation to complex decision support system using artificial intelligence technologies. The choice of appropriate technology is an involved task, because it requires mapping customer requirements to technical features and benchmarking and comparison of disparate technologies.

Make Sure They Come Back: Just when you thought that the technical challenges are over, you realize that the next hurdle may be even higher. In most service organizations, technical support knowledge is always evolving and is never complete. As the initial excitement following a successful technical implementation subsides and the development team finds new opportunities to pursue, the self-service knowledge base often remains unattended. If the content is not continuously updated with new product and support information, its efficacy deteriorates, leading to value attrition and eventually to user defection.

About Joe Barkai
Joe Barkai is the founder and principal of Diagnostic Strategies, a management and technology consulting practice specializing in Service Lifecycle Management (SLM), Diagnostic Knowledge Management, Service Process Optimization, and Advanced Diagnostic Solutions. Joe brings over 15 years of consulting experience to Fortune and Global 1000 companies in the US, Canada, France, Germany, Finland, Italy and Israel. His business consulting practice covers service business strategy, service process reengineering, and implementing support automation technologies, diagnostic expert systems, and advanced diagnostic and prognostics methods.

About Diagnostic Strategies
Diagnostic Strategies is a research, advisory and consulting firm focusing on Service Lifecycle Management (SLM) practices and technologies. Diagnostic Strategies works with Fortune 1000 companies on service business process and strategy development and implementing advanced solutions that include self-service, expert systems, remote diagnostics and prognostics in their service operations.