Enterprise Search is Maturing
flickr.com/sean_oneill
When we started Comperio 6 years ago, many of the search projects were limited to one or two objectives. The goal was often just to “fix” search on the web site and/or the intranet. With the tools and solution concepts we had available at the time that was demanding enough.
Today, however, the success of the first wave of search solutions have led our customers to ask for “more search”, “more access” and “new insights” into their vast amount of information repositories. While the idea is not new, we are now seeing innovative and true new enterprise-wide indexes or information marketplaces being established. Structured data is at the core and the solutions are more centered towards Information Access than the previous search-only focused solutions. These solutions include every digital content and data repository that the enterprise possesses and even additional external sources that potentially bring new insights. The use-cases and business purposes has become more comprehensive and thus requires a more flexible way to service customer-, partner-, authority- and/or employee-facing applications.
Enterprise Search is maturing and that’s good news: think of all the new opportunities this opens!
Nice article, Jørn!
A good description of how our business has changed over the years – for the better. I recognize the shift from plain bolt-on search to universal enterprise-wide information access. I hope it means that decision makers are comming to terms with two inconvenient truths about great enterprise search: 1) it demands a high initial investement up-front, as well as 2) significant and continous maintenance costs to keep both the index and the user interface sharp. Magic does not come for free :-)
Well observed Jørn.
My experience working with customers is that as soon as the internal “information marketplace” is established in the form of an enterprise-wide index, a whole bunch of new and innovating IT solutions are made possible. These solutions typically is developed as new views into the same index based on the previously established information model. Let the imagination prevail!
I’d go even further and say that search technology is heading to become tomorrow’s integration layer and a place for business process modeling. The concept of “actionable result sets”, which we see and implement everywhere these days, is the first baby steps in that direction.
An enterprise search solution is always put in a central position, right in between source systems and end-users. There is by definition, and necessity, good connectivity in both directions. Scalability and fault-tolerance is already catered for. And most importantly, truly schema agnostic indexes are no longer just a farfetched idea. I think it’s safe to say that slicing and dicing of enterprise content, disregarding in which data repository the actual content belongs, is just around the corner!
Perhaps “corporate mash-up” is the next big buzz word? I said it first. ;)
And to follow up on “truly schema agnostic indexes”. See this subproject being done at NTNU for the iAD Research Centre on Information Access Disruptions:
http://www.idi.ntnu.no/grupper/db/research/projects/ongoing/iad.php