Beyond Search

How often have you called a company for help with a product you bought from them, only to be frustrated by their inability to give you the correct answer quickly?

Most enterprises struggle with the challenge of interactive content "findability." When contact center agents search for content, they get too many search results or too few. As a result, most agents develop their own proprietary cache of tips and tricks to serve customers. Unfortunately, customer and product issues change over time. Agents cannot keep up without technological help, hence most contact centers face broad agent-productivity spreads (read higher cost and poor customer experience) .Why doesn't search, something that works so well on public Web sites, work as well for customer service? And what is the solution?

Search is not a panacea

Search is a powerful tool-cost-effective, easy to deploy, and universally usable. Further, the success of search on the Web has made it a popular tool in contact centers. But there are several reasons why search does not work well in contact centers, most of which have to do with differences between the public Web and Intranets.

Customer expectation difference

Public Web users typically seek one of many correct or nearly-correct answers. Most of the time, they do not look for an exact document because they are simply researching and browsing-a phase of customer engagement when directionally correct results suffice. If, in fact, they are looking for an exact document, they are willing to sort through approximate results, based on search. They don't expect much because Web searches are free. So they are happy with what they get.

The situation is completely different when customers call a contact center. Typically, they have bought the company's product, so they expect fast, accurate resolution. They know what their problem is and they expect a quick, correct answer. This difference in customer expectation is the biggest reason why search works "acceptably well" for Web search but not "at the moment of truth," when customers seek specific information on a business Web site or from the contact center.

Inherent structure in Web content

Unlike Intranet content, public Web is intrinsically structured through hyperlinks. Every Web page links to other public Web pages. Search engines rely heavily on well-known hyperlink processing algorithms to extract inherent structure and relative importance on Web pages. On the contrary, Intranet content has minimal hyperlinks that can be similarly exploited by search engines. Therefore intranet searches tend to return hits that are not ranked as smartly as public Web searches.

Hidden investment in structuring Web content

Finally, the secret of public Web search! Google and Yahoo! invest heavily (and quietly) into analyzing typical search terms, and annotating the content to return "voodoo quality" responses to search queries. Intranet search engines do not benefit from such investment.

Clearly, the solution to finding the right content in contact centers or in customer self-serve systems has to go beyond search.

Multimodal Access (MMA) delivers on findability

MMA is an approach to finding content that goes beyond search to include the following access options:

• Keyword and NLP search;
• Topic based browsing;
• Contextual FAQ;
• Virtual Assistant; and
• Guided Help.

MMA matches the access method to multiple factors, including:

• Inquiry type;
• Customer expectation;
• Agent expertise; and
• Content capability.

For instance, if the customer is asking a common question, it makes sense to direct them to an FAQ page based on NLP or keyword search. But if the customer has a diagnostic or advisory question, guided help access works better.

Business case for MMA

Depending on the type of access method deployed, costs and benefits vary. Search typically does not involve any ongoing content management or structuring costs, whereas guided help may require knowledge authors who review/refine case bases to guide the agent or customer to the right content. The relatively higher cost of guided help is more than justified by its significant benefits for appropriate use cases. Typical payback period is less than one year-based on reduced escalations, training costs, and improved first contact resolution (FCR).

Further, MMA's flexibility can be leveraged across different business scenarios-front office, back office, and Web self-service. This flexibility minimizes TCO through standardization. It also improves business agility because the learning curve benefits of MMA-especially when delivered on a single software platform-help businesses rapidly respond to evolving requirements.

About the Authors

Ashutosh Roy is chairman and chief executive officer for eGain Communications. Prior to eGain, he cofounded and served as chairman of WhoWhere?, which is now part of Lycos. Mr. Roy also cofounded Parsec Technologies and worked at Digital Equipment Corporation. For additional information, visit www.eGain.com.

Anand Subramaniam is the vice president of worldwide marketing for eGain. Prior to eGain, he served in executive, managerial, and tactical roles in marketing, sales support and product management at companies such as Oracle, Lotus, Intel, and Autodesk, as well as startups.

Please note that the Viewpoints listed in CRM magazine and appearing on destinationCRM.com represent the perspective of the authors, and not necessarily those of the magazine or its editors. If you would like to submit a Viewpoint for consideration on a topic related to customer relationship management, please email viewpoints@destinationCRM.com.