Enterprise search platforms are fairly easy to understand. A user types in some keywords and the engine returns (hopefully) relevant documents, emails and other corporate content. Search-based applications, on the other hand, are a little more subtle.
There are an almost infinite number of ways to use enterprise search software – and frequently, “it’s being used without people even realizing it,” said Whit Andrews, an analyst with Stamford, Conn.-based Gartner Inc.
Search-based applications are software programs whose search function is powered under the covers by enterprise search technology. In fact, the familiar search box is often not even seen, and end users don't have to write their own queries based on strings of keywords – the primary benefit of such applications over traditional enterprise search engines.
Instead, users go about their daily business. As they do so, the information they access and interact with in business applications is made available through complex search queries generated by the enterprise search tools at work in the background.
Seth Grimes, founder of Alta Plana Corp., a Takoma Park, Md.-based consulting firm that follows the enterprise search market, cited a potential voice-recognition application.
The goal of search-based applications: finding info without being asked
“You might be a call center representative talking to someone, and a particular product comes up in the conversation,” Grimes said. “A search-based application would bring up that product information, or it would bring up transaction information automatically without having to have someone type information in a box and hit the search button.”
Users don't even have to know what they're looking for in many cases. For example, with a search-based inventory application, a worker trying to find a warranty document for a coffee maker might use a navigation tree – clicking on products, then kitchen products, then coffee makers, and so on – until he locates the document.
The navigation tree isn’t a static entity, though. Searches are still happening, but the application is constructing the queries necessary to find the desired information based on the twigs of the content tree. And because the application itself is writing the queries, the results often are more precise than users could get on their own.
“What's happening in the background is the application is submitting a query to the search index saying, 'Give me all the products which are in this category,'” said Shawn Shell, a consultant with Chicago-based Consejo Inc. “The construction of that particular query can be quite complicated because it has to consider a number of different metadata value combinations to arrive at a very, very relevant and specific results set.”
Possible uses for search-based applications
As Andrews said, there are a variety of potential search-based applications. Currently, the more popular include applications used in call centers to help customer service representatives find relevant content they need to assist a customer while still on the phone – the kind of use that Grimes described. There are also some business intelligence (BI) platforms that rely on sophisticated search algorithms to produce custom charts, graphs and other data visualizations.
But enterprise software vendors still have a long way to go in integrating enterprise search into new and different types of applications, such as customer relationship management and enterprise resource planning software, Grimes said. “We need to make search more effective, and one way to do that is the integration of search with forms of business intelligence, and further integration of search with business processes where search goes on behind the scenes.”
Unfortunately, organizations are really at the mercy of vendors when it comes to developing search-based applications. Such programs are simply too complex for most organizations to build internally, Grimes said.
The good news is that the opportunities to leverage enterprise search tools to improve existing applications are nearly limitless. The potential improvements that better search capabilities could bring to BI and data analytics software alone are formidable, analysts said.
The key to effective search-based applications, though, is for the enterprise search process to do its work in the shadows, Grimes noted. “Search just has to be something that goes on while you go about your business,” he said.