Nine Reasons Why Enterprise Search Is Sinking

September 30, 2008

I’ve been thinking about the distribution of attendees at a recent combo trade show I attended. I won’t name the show, and instead I want to focus on summarizing the comments made to me by attendees and exhibitors. I’ve chosen to present this list in terms of enterprise search but it applies to content management and knowledge management as well.

  1. When people say “search”, the people are not talking about key word and Boolean queries. Search is in trouble, therefore, because of the assumptions about the meaning of the term. Just as in high school debate, definition of terms is the first step. Get this wrong and you have the craters that pock mark the enterprise landscape.
  2. Search looks easy but is not. I’m not sure if we can pin this on Google. The company uses kindergarten colors and a very simple interface. The user thinks, “Wow, this is easy.” Search in the organization is rocket science, yet no one believes it until projects become problems and budgets run wild. Misjudge complexity and you pay, and pay big.
  3. Vendors make sales. Vendors are not college professors who educate you. As a result, I hear such statements as “The vendor told me x, y, and z.” Well, what do you expect. Search sales professionals move around more than the Dallas Cowboys defensive secondary. The vendors want to generate revenues, not experts in search. So, customers have to know about search, ask the appropriate questions, and know when goose feathers or giblets are served up.
  4. Internal information technology wizards assume that search is a no brainer. As a result, the internal wizards nap during training or skip it, ignore advice of consultants, and dismiss cautions from the vendors’ engineers about infrastructure. When the search system fires up, the system falls over. The IT professionals say, “Hey, not my problem.” I beg to differ. IT professionals need to learn to listen. If most of these engineers were Google or Microsoft R&D grade, the engineers would be–you know what’s coming–working at Google or Microsoft or involved in rocket science. My thought, “Skip the genius pitch and learn how to make search work before deploying the system.”
    image
  5. Users. Accept this: Users cannot accurately and completely tell you what they need and will use for information access. You have to conduct research, offer demonstrations, and provide proofs-of-concept. Skip these steps and the most influential person in the organization will hear from Timmy or Becky that “Our search system still does not meet out needs.” Research, communication, and involvement are essential. Skip these steps and you will fall off the tight rope into the abyss; that is, you will sit in your cube and the email chime won’t sound. You are excluded.

  1. Lawyers. Most search systems are created without a thought about litigation and eDiscovery. Not in today’s world can you move forward ignoring legal eagles. Once a matter is “real”, search has to deliver or be excluded. Have you tuned your system to support eDiscovery? If not, check out the openings at the local Burger King or Wal*Mart. Lawyers are not forgiving when it comes to search and FRCP.
  2. Marketing. Let’s be direct with one another. Search does not mean searching trade secrets, knowing which employee has a dependency problem, or having the salaries of the regional VPs. Search applies to general content, and one of the squeakiest wheels is marketing. Marketing wants to know what it needs to know because the department is in their eyes really special. You will have a headache unless you get marketing on board and keep its entitlement-trained experts happy.
  3. Consultants. After walking around this last show, I saw more consultants than search and content processing vendors. There were two groups whom I determined with the technical resources to make a flawed search or content processing system work. Following the directions and advice of a former marketer, indexer, junior executive, reference desk specialist, or journalist is an invitation to disaster–fast.
  4. Show organizers. I won’t mention these professionals, but in the last three years, I have concluded that the texture and content of shows. If you see the same vendors and hear the same talks (even if given by different people), the likelihood of making a break through drops off. With attendance at search shows declining world wide (based on my appearances so far this year), you have to work quite hard to find the hot, new, leap froggy systems. Gold fish don’t comprehend rain because the little beasties can’t escape their environment.

Let me know if I am on the beam or just flapping my wings and honking. Use the comments section to this Web log. Stephen Arnold, September 30, 2008

Comments

3 Responses to “Nine Reasons Why Enterprise Search Is Sinking”

  1. Why Enterprise Search Is Sinking on September 30th, 2008 8:07 am

    […] Article about enterprise search marketing from Stephen Arnold on his experiences on a lately visited trade show. Share and […]

  2. anonymous on October 6th, 2008 8:42 am

    There may be a tenth reason.
    The system(s) work(s) well-enough and all the dots we”ll ever need are there. But the right dots rarely get connected even in the best systems. So the dots and the infrastructure carrying them get blamed.Search sucks because it doesn’t replace perception and imagination (and whatever else it takes). Can we figure out the DNA of that? Can Intel put it on a chip? And then will all we have to do is ask? Isn’t that what we do now? How would all of this work if it could work? And if it could work, would it matter? I doubt it, but if anyone would like to sponsor me, I will look.

  3. MarkLogic 4.0: A Next-Generation Content System : Beyond Search on October 7th, 2008 12:05 am

    […] the nine problems that most search and content processing companies struggle to overcome. Click here for these […]

  • Archives

  • Recent Posts

  • Meta