Impact of enterprise search on customer service

I recently researched the value of enterprise search technologies on the customer service function with my fellow analyst Sumair Dutta.  Search and customer service are not exactly strangers as 44% of Aberdeen’s survey respondents have been using search in that role for almost 5 years.  But, what is changing is the approach that is being taken.  Organizations are finding it increasingly difficult to provide contact center agents with the information they need in the time required.  In Aberdeen's 2011 customer service contact center research (The Contact Center in a Profit Centric Service Organization) companies rated themselves an average 3.4 (on a 1 to 5 scale with 1 being poor and 5 being excellent) on their ability to provide contact center agents with access to information. Further, organizations noted that more than 20% of agent time was actually used in looking for resolution information. This extends overall resolution times, reduces first-call resolution rates, and adds to the costs of the service organization.  From a revenue generation viewpoint, it also takes time away time for up-selling and cross-selling.
 
However, where enterprise search has been used to support customer service there has been a strong return in customer retention and subsequent revenue driven as a result of improved retention (Figure 1).

One of the foundations for these gains is a distinct improvement in the first time resolution rate within the service center, to the tune of 8% for organizations that leverage enterprise search capabilities. This was accomplished primarily by ensuring that support and service teams are able to quickly find the information they need to help customers resolve issues, while also allowing support agents to educate customers about complementary products and services.
 
Companies have put in place several capabilities and characteristics to deliver this level of performance in the service centre.  One overriding theme stands out though – the recognition that search is not a generic commodity product.  In other words, search isn’t something that you can just drop in and leave – at least not if you want to extract real value from your investment.  There are a couple of deliberate steps that businesses can take to ensure that their service centres get the most value from their search technology of choice. 

These are:

> Ensuring that business users are involved in defining the requirements for search.  Just over half of all organizations currently do this, but an additional 32% plan to take this step in the next year.  I see this as an acknowledgment that treating search as a commodity doesn’t deliver results.  Like any other piece of sophisticated software it needs care and nurturing for it to take root and flourish.  In that same vein, similar numbers of companies also have an explicit mechanism to capture feedback on search performance.  That is, they provide a pop-up box, or a box on the page of search results allowing a user to provide feedback on the relevance – or otherwise – of search results.  This feedback step then becomes one of the inputs into the roadmap that defines the future evolution of the search solution.

> Seventy-six percent (76%) of survey respondents are working towards a role-based search implementation.  This is another indication that not all search is created equal and that different people in different job roles need different things from search.  The search needs of someone working in a customer service center are very different from the needs of someone working in human resources. For one thing, the information that the two users are searching for is going to be very different. Human resources professionals are unlikely to require access to detailed internal documentation on products and services, for instance. Likewise, knowledge workers in a service center are often working on time-critical problems where finding the best solutions to a customers problems at the first time of asking can reap huge dividends.
 
Beyond these organizational steps, service centres also look to capitalize on key technology features such as filters and clustering.  Filters based on tags and metadata can help to quickly find the most appropriate search results. For example, filters might be used to choose only results that were related to the particular product that the customer was using. In parallel with this, over 90% of organizations plan to include document attributes in the search results. Attributes, such as file type, also help service staff quickly find the most appropriate documents to provide timely assistance.  Clustering search results can provide similar guidance. Clustering dynamically groups search results into categories of related information. By creating clusters "on-the-fly" a more flexible grouping of documents is possible than would be provided by filters alone which, by definition, need to be pre-defined. Although clustering is used by a minority of customer service organizations currently, 47% plan to adopt this approach within the next year.
 
David White, Senior Research Analyst, Aberdeen Group.