Enterprise Search engines can be divided neatly into two categories: those optimized for website search and those optimized for searching across internal information silos. Today the gap between the two is opening ever wider.
The reasons are not too difficult to understand. External websites that feature customer interaction are considered a priority, especially for ecommerce environments. If your customers can't find what they're looking for, then that is bad for business.
To be sure, searching and providing navigation to an external website is not usually cheap or easy either. But it has one major advantage over internal search: the data wants to be found, and is typically structured, stored, and tagged accordingly. Most of us are familiar with the very granular and typically very accurate faceted search provided on today's large shopping sites. Search in this environment works well and there is a growing market for it.
Internally focused Enterprise Search remains, as the awful phrase goes, the poor stepchild. Searching multiple internal silos -- full of unmanaged and unstructured information -- is typically a hard, expensive, and disappointing task to undertake.
So guess where all the Enterprise Search vendors want to focus their efforts these days?
You can't really blame them of course, not the least because the needs of ecommerce and external websites extends far beyond Search. Being able to find a replacement fridge drawer on the Samsung website (as my wife did today) is scratching the surface of what could be done. As Mike Davis of Ovum said at the recent European Enterprise Search Summit, "Firms like Tesco drive their business from the data on your loyalty card, but they want to know more about you than your transactions." Ultimately its all about context. As we have said many times before, the context for structured data is often found in an unassociated unstructured file.
And so the world of Search enters the world of true analytics and "Big Data." What has long been the sole purview of Business Intelligence vendors is now slowly starting to be encroached upon by Search tools from IBM, Oracle (Endeca), HP (Autonomy) and a few independents such as Coveo. I imagine we'll still see two different takes at the same problem for a while, but now that Big Data organizations like IBM, HP, and Oracle are taking Search seriously for once, over time some kind of solid hybrid may just emerge.
Customer interaction and commerce will grow ever more sophisticated, with predictive analytics taking the lead. No doubt we will see the split between Internal and External Search widen even further over the next couple of years, as some Search vendors at least have finally found a truly lucrative niche, and they are unlikely to turn back now.