New Enterprise Search Market Study

August 1, 2017

Don Quixote and Solving Death: No Problem, Amigo

I read “Global Enterprise Search Market 2017-2022.” I was surprised that a consulting firms would invest time and energy in writing about a market sector which has not been thriving. Now don’t start sending me email about my lack of cheerfulness about enterprise search. The sector is thriving, but it is doing so with approaches that are disguised as applications which deliver something other than inflated expectations, business closures, and lawsuits.

Image result for don quixote

I will slay the beast that is enterprise search. “Hold still, you knave!”

First, let’s look at what the report covers, then I will tackle some of the issues about which I think as the author of the Enterprise Search Report and a number of search-related articles and analyses. (The articles are available from the estimable Information Today Web site, and the free analyses may be located at

The write up told me that enterprise search boils down to these companies:

Coveo Corp
Dassault Systemes
IBM Corp

Coveo is a fork of Copernic. Yep, it’s a proprietary system which originally was focused on providing search for Microsoft. Now the company has spread its wings to include a raft of functions which range from the cloud to customer support / help desk services.

Dassault Systèmes is the owner of Exalead. Since the acquisition, Exalead as a brand has faded. The desktop search system was killed, and its proprietary technology lives on mostly as a replacement for Dassault’s internal search system which was based on Autonomy. Most of the search wizards have left, but the Exalead technology was good before Dassault learned that selling search was indeed a challenge.

IBM offers a number of products which include open source Lucene, acquired technology like Vivisimo’s clustering engine, and home brew code from its IBM wizards. (Did you  know that the precursor of PageRank was an IBM “invention”?) The key is that IBM uses search to sell services which have a higher margins than providing a free version of brute force information access.

Microsoft is interesting. For this short blog post, I won’t drag up the in house innovations, Microsoft’s numerous search systems, or its purchase of nCompass, Jellyfish, and other “findability” systems. Microsoft bought the fascinating Fast Search & Transfer company and technology. The fact that Fast Search experienced some legal and financial headwinds is important to me. Fast Search overpromised and apparently over reported some of its earnings when its repositioning into an enterprise search only vendor provided to be a less than rousing success. Anyway, Fast Search seems to chug along within SharePoint at least.

Oracle owns a number of search systems. These include the Applied Linguistics system, TripleHop, Endeca, among others, as well as the Secure Enterprise Search system. Oracle appears to have demonstrated that secure search does not sell. What sells is the basic search function available within the Oracle database.

SAP now pitches softballs with NetWeaver. The company’s cleverly named TREX is a goner I believe. NetWeaver is described in a brief blog post at this link. Other NetWeaver links are timing out as I assemble this blog post. SAP is in my view not a vendor of a separate enterprise search system like Coveo, for example. When a big company bites into the juicy SAP enterprise Swiss Army knife, NetWeaver is the search system. (NetWeaver’s origin is anchored in 2001 and that document is still online as August 1, 2017 at this link unlike many other SAP NetWeaver links.)

What is interesting about this list is what it omits. Now I don’t think some of the following systems are able to handle the challenge of rich media, social media, assorted file formats, and the unstructured data which floods organizations. But for many organizations, these outfits offer a reasonable way to locate information in an organization and they should be discussed, not ignored:

  • Antidot
  • Attivio
  • dtSearch
  • Elasticsearch from Elastic and its resellers’ solutions; e.g., SearchBlox
  • Fabasoft Mindbreeze
  • Funnelback
  • Maxxcat
  • Rocket (Aerotext)
  • Sinequa
  • Leidos (Teratext)

What I find interesting is that there are others and many newcomers, acquirers, and MBA-inspired optimists chasing what is now a 50-year-old dream: High precision, high recall, low cost, ease of use, minimal engineering and customer support demands, scalable, speedy, up-to-the-second indexes, and user pleasing search and retrieval. In a sense, creating an enterprise search system with Google’s special project to solve death. It’s a tough nut to crack.

The big story in search is the demise of proprietary systems and the surge in the open source options, particularly Elastic and its Lucene-derived Elasticsearch. There are other important vendors who provide “search” but these are firms which have carefully avoided being dipped in the tar and feathers of old school information access; for example, BAE Systems NetReveal (Detica) and Palantir Gotham, among others.

The report, therefore, strikes me as skewed to a handful of vendors who presumably enjoy surging revenues and profits from their sales and services businesses. Well, that’s one hypothesis I suppose.

My other concern with the approach to enterprise search is the notion of “clauses.” It seems from the write up referenced above that a “clause” is a statement of belief, hope, business school chopped liver, or a convenient categorization method.

Here are three example clauses:

  • Clause 2 and 3 studies the key Enterprise Search market competitors, their sales volume, market profits and price of Enterprise Search in 2016 and 2017
  • Clause 7, conducts the region-wise study of the global Enterprise Search market based on the sales ratio in each region and market share from 2012 to 2017
  • Clause 11, 12 and 13 present the competitive situation among the top manufacturers, with sales, revenue and global Enterprise Search market share in 2016 and 2017;

These would be interesting data. But in the work I have done and continue to do on a modest scale, the data are not available. The companies either don’t reveal audited figures or break out their financials to make it possible to get reasonably accurate sales and profit data.

The reason is simple. Search is expensive and often a money losing proposition for its stakeholders. Think of the struggles Hewlett Packard has had with Autonomy. Consider the Fast Search financial allegations and eventual jailing of an executive for financial issues. Ponder the history of Attensity, Convera, Delphes, Entopia, Siderean Software, and many other enterprise search vendors. Think about Google’s termination of its search appliance and getting back to good old online ad sales for real money.

Perhaps the data in this report are stellar with a great analysis tossed in to boot? For me the net net: If this enterprise search study can breathe some new life into what strikes me as a low growth, behind the eight balls enterprise software sector, great. For now, go with a hosted Elasticsearch implementation and start shopping for a next generation information access systems. I identity a number of these in my 2015 study CyberOSINT.

The point my research boiled into a flinty residue was:

One sells search by not using the word search.

This write up happily uses the phrase “enterprise search.” That’s a bit of a problem from my point of view.

Stephen E Arnold, August 2, 2017


One Response to “New Enterprise Search Market Study”

  1. on August 2nd, 2017 8:09 am

    New Enterprise Search Market Study : Stephen E. Arnold @ Beyond Search

  • Archives

  • Recent Posts

  • Meta