Start Up Flops: No Search Vendor Examples

January 21, 2014

I read a quite interesting article, “51 Startup Failure Post-Mortems.” The links in the story point to source documents. The article presents a summary of the principal take away from a crash-and-burn experience.

The lessons are fascinating and include:

  • Spreadsheet fever; that is, making assumptions about how many customers a start up will capture and how much each customer will pay
  • Users don’t want what the company is selling
  • Misallocating time; that is, buying stuff, not selling stuff
  • Management errors; that is, hiring, spending, marketing.

I recommend this article. I want to point out that none of the examples came from the deep pool of search-and-content processing failures. Outfits involved in information retrieval face some hurdles that put some spin on the problems outlined in 51 Startup Failure Post-Mortems.

Let me highlight three:

First, search is assumed to be over with Google the winner. Furthermore, most individuals assume that their search skills are quite good, even excellent. As a result, a new information retrieval vendor has to find a way to capture the potential customers’ interest. Indifference combined with a perception that the potential customer has good enough tools produces unconscious resistance to a new search system. Search is ubiquitous and works reasonably well for finding a phone number, a pizza joint, or news about a celebrity.

Second, search and content processing has a cash appetite that has continued to surprise innovators, entrepreneurs, and acquirers. As digital information becomes more plentiful, more resources are required to process the information. Search is expensive and over time only gets more costly. Even the “winner” Google is taking steps to control its Web search costs. Money is a problem for search vendors from Day One and continues to be a problem for the lifetime of the search vendor. Exponential information growth requires exponential investment in resources.

Third, search does not work. Sure, on the surface, a quick look at Bing, Google, or Yandex provides results that answer most  users’ questions. However, when one tries to dig into a subject, search in its present form is often unwieldy if not broken. For example, research a topic through time and one finds that content is no longer available. Commercial databases are fraught with gaps. Free services skip content due to latency problems or a desire to process certain low hanging fruit. In short, finding information today is getting more difficult by the day. Online content is not comprehensive, but users assume search systems have “everything.” Wrong.

These three characteristics contribute to the quiet and often noisy failures of search and content processing systems. Search is so disappointing that marketers find it easy to over promise and the systems then under deliver. Some of the lessons about actual companies that have failed in the search sector appear in the Xenky profiles.

Perhaps the next version of “51 Startup Failure Post-Mortems” will tap into the search and content processing sector.

Stephen E Arnold, January 21, 2014


Comments are closed.

  • Archives

  • Recent Posts

  • Meta