The Enterprise Search "Essay Test"

Andy Moore: There are plenty—I’d say MOST—businesses which believe that a "merely OK" search tool that comes more or less free with their infrastructure is "good enough." Is "good enough" ever good enough?

Jason Hekl: In situations like that, I think you call into question the primary objectives of that purchase decision. When an IT organization is driving the decision, search is often considered a commodity feature, and infrastructure maintenance and cost concerns trump all other objectives. When we face that situation, we encourage the people involved to articulate their vision of how that investment choice will allow them to create a differentiated and compelling user experience for their customers, employees and partners. "Good enough" just doesn’t cut it when the objective is to create value through a better customer experience.

One of our customers is implementing InQuira as part of an overall brand loyalty strategy. This wireless provider analyzed what it would take to achieve the type of emotional and functional loyalty enjoyed by the likes of Apple, Prius and Whole Foods, and determined that the online customer experience was essential. In their words, "Establishing loyalty and trust via our interactions can establish a precedent in our laggard industry. This can be done through initiatives that change the customer engagement." One of the key mechanisms for engagement in this improved online customer experience? Search. They view it as a loyalty-building differentiator.

John McCormick: SharePoint is indeed good enough for certain solutions and business problems, but, obviously, it is not as powerful as is needed for many situations, hence the recent Microsoft acquisition of FAST. Bottom-line: you have to be careful with "good enough." How many times have you seen a company buy a solution that was "good enough," only to later buy yet another solution that was also deemed "good enough" for another problem du jour. Or, that initial technology is replaced by a more comprehensive solution. The "good enough" line of thinking often serves short-term needs, but can create mid- to long-term havoc for the enterprise. Best advice: take a step back, consider and gather your requirements carefully, and go for a solution that can address your immediate pain, while also being flexible and scalable enough to grow with your business.

Jerome Pesenti: This is one of the main reasons why many "good enough" search engines deployed in organizations today end up not being used for very many searches. What IT sees as "good enough" turns out to not be good enough for end-users, abandoned after one too many fruitless searches. Historically, when Vivisimo replaces one of these "OK search tools," our customers report a dramatic increase in search usage even during pilot periods when the only publicity for the new search is word of mouth.

The challenge for search is that it needs to be both ubiquitous and universal while respecting specificities. Companies have responded to this challenge by either deploying point solutions—which can be efficient but see little usage because they silo the information and need to be found before being used (which is ironic for a search engine…)—or deploy one-size-fits-all solutions which treat all content and all users in the same way. These are not "good enough" solutions unless the definition of "good enough" means that workers still can’t easily find and share information.

A good solution is one that can address: both the universality and specificity requirements at the same time; be both distributed (from an resources/admin standpoint) and centralized (from a user standpoint); can leverage the specificities of each type of content while being consistent across all content; and takes into account the specific needs of each department and each user while offering sensible default behavior.

Think 1998, and remember the advent of Web search. When Web search became truly "good enough," it started a revolution. Ten years later, search still has made very little headway inside the firewall, but things are about to change.

Vijay Koduri: Many businesses in the past few years have been searching for infrastructure that can help their employees share information better across the world. However, effective sharing really has two parts: "uploading" and "finding." Most content tools focus on the first part (uploading) by creating central repositories. What they don’t realize is when you populate this "central repository" with hundreds of thousands of documents, the second part (finding) becomes next to impossible without adequate search capabilities.

Findability trumps the ability to create a central repository. And true findability can only be achieved through a world-class enterprise search engine with the highest relevancy.

James Waters: If you are settling for "merely OK" search, you’re not aware of the time and money and aggravation this is causing your knowledge workers. Most applications contain a light search, but I guarantee you, the users are not happy with it. And that translates directly to the people they service. Wrong or incomplete answers translate to bad customer satisfaction scores = customer turnover.

Harald Jellum: "Good enough" is a subjective view. The horse-drawn carriage was good enough for many, even in the days of Henry Ford’s T-model. The same goes for low-end search technology. We have yet to come across customers who do not see a value in search. There is, however, some price sensitivity dependent on the customer’s intended use.

The challenge for search technology vendors is addressing business issues and communicating the value-add to customers. Our experience shows that the larger the organization, the larger the need and the business potential across a number of functional areas. The untapping of the business potential of search is not a question of implementing a simplistic and "good enough" search tool, but understanding and addressing how the technology can support their knowledge workers and business processes. This requires search and monitoring functional and technical expertise that customers need in order to create best-in-class processes. We expect that we have just seen the start of this process—what is viewed as "good enough" by some today will in a few years be as outdated as the horse-drawn carriage.

Johannes Scholtes: This is absolutely true. Another consideration needs to be mentioned: the broad availability of free personal and open-source enterprise search tools that can perform an array of serviceable tasks for many users. Buyers and vendors know that enterprise search platform providers are going to struggle in a market where the good-enough part of many solutions is free.

Think back to the enormous impact the Windows NT free search engine had when it was introduced in 1993. Many vendors have gone out of business since then, except for the ones that adapted to the changing marketplace.

Vendors that have survived and will continue to survive must operate in specific technological niches or focus on search-driven applications in which specialized search capabilities are integrated within an applied business activity (contract management, e-discovery and e-disclosure, litigation support, e-mail management for compliance, historical archives, correspondence management in litigious environments and so on).