After giving a lecture at the Fall 2013 Enterprise Search Summit in Washington, DC a VP at Giant Media Company asked me the question: “What’s ahead for enterprise search in 2014?”
As I began relaying the canned response that I usually save for these “predict the future” queries, to my surprise, several people who were wandering past my grouplet stopped to listen. Despite the somewhat autumnal talks I heard, the interest in a nearly dead software sector surprised me and spawned several questions: Why is there such an interest in enterprise search? Why are organizations still struggling to find documents after 50 years of “innovation” and hundreds of millions in investment in a utility service? Why are conference attendees, after three days of commentary, still struggling with what seems to me to be a very basic question?
Here is my attempt to answer some of these questions. Namely, what is in store for 2014? Over the next year, there will be more attrition among search vendors. Some outfits will just fade away. (No, I won’t mention any of the firms on life support. Executives at these firms regale me with tales of resurgence. I don’t have the appetite for this type of rah rah. Microsoft, take note, I am not saying Fast Search & Transfer is a dead duck.) Attrition, as I use the term, includes acquisitions. There will be some snap ups, but these deals will lead to the absorption of the “brand” into the acquirer’s product line. This is largely the fate of Autonomy, Brainware, Endeca, Exalead, InQuira, ISYS Search Software, RightNow, Vivisimo, and any firms I have overlooked. The gap or void in proprietary enterprise search will continue to be filled by quasi proprietary systems like ElasticSearch, LucidWorks, Searchdaimon, Sphinx Search, and the many other open source services available. There are some “pretenders” to the revenue glory of Autonomy before it was acquired by Hewlett Packard. But in reality, these vendors lack brand clout and have to amp up their marketing to have a shot at hundreds of millions in revenue. Net net: Lots of buying, failing, and start upping.
Second, the technology to distribute search functions across multiple, lower cost computing resources will expand. Most vendors at least talk about a cloud option. Some like Searchdaimon can deliver. Unfortunately most of the other vendors who yammer about massively parallel, distributed, scalable, extensible cloud solutions are scattering fairy dust. Cloud solutions are easier for vendors to support. Lots of customers use the vendor’s platform. However, for many organizations, the cloud is a bit of a problem. There are pesky contractual requirements for some government work. There are costs that depend on old fashioned taxi meter pricing tables that “kick in” when certain thresholds are crossed. There is the unfortunate issue of latency at different points in the information retrieval system that make some Fancy Dan solutions run erratically or slowly. Net net: Technical advances in infrastructure will outpace innovations in search technology.
Third, interfaces are now positioned as a way to improve information retrieval. Sure, interfaces are a big deal for those who want to get General Mills’ type information. You pick a box and the cereal company decides what goes in it just like Google’s mobile search. Next year interface will be an even bigger deal. The reason is that interface distracts the child in the enterprise as a shiny watch mesmerizes a baby. The hard problems like what content is available, what specific units require what specific information complete a work task, what content is accurate, and what content is needed to answer a business question are ignored. In 2014 it will be a lot more fun to go to a meeting to talk about colors and icons. Net net: The visual revolution will push tough questions out of sight.
Finally, mobile usage will make personalization an essential. With that personalization, filtering becomes the go-to function. The future of point-and-click means that more workers will be taking what the system gives them. If the system provides the right information at the right time, life is good. If the system provides flawed or off point data, maybe the lousy decision making will accelerate. Exciting. Net net: Mobile almost guarantees a continued devaluation of nuts-and-bolts search and retrieval.
In regard to the interest in enterprise search that followed my talk:
I think some of the interest is more in job getting and job keeping than search. Some of the folks in the hallway grouplet showed heightened attention when I pointed out that folks get fired when the “new” search system disappoints users or the organization’s chief financial officer.
Additionally, I have mounting evidence that the confusion about “which system” to select is only escalating. The reason is that the available proprietary systems have to be compared to and contrasted with the “free” open source search systems. Believe me, the procurement job is getting harder. That, of course, feeds risk and stokes anxiety. At this time, none of the mid tier, azure chip consultants are much help. “Industry leader” IDC bought some information from me and did not pay me for my work. The vaunted IDC “search mavens” seem more confused than the deer in the hallway grouplet. Third, money is tight. Search is viewed as a service that has bitten some companies in the ankle before is now viewed with great skepticism. As a result, the individuals who have to “fix up” the incumbent system or license a “next generation” system are not getting much, if any, management support. The result is dithering. There you have it, the outlook for 2014: Uncertain buyers, revenue hungry (maybe desperate vendors), annoyed system users, and unconvinced senior managers. Search is interesting. Enterprise search will remain a digital turkey. The consultants will almost certainly give thanks.