It's remarkable to me that, at this late date, a meaningful number of publishers continue to jeopardize their futures with a less-than-total commitment to developing Web-based product offerings.Too often, I hear from publishers, "yeah, we're on the Web," delivered with vocal inflection that makes it clear the Web is viewed as some annoying new obligation, instead of the future of their businesses.

Of course, when you view something as an annoying obligation, you do what's required of you as quickly and as cheaply as possible.

Not surprisingly, this attitude reflects in the finished product. More than once, I've been told by programmers that they have been handed databases with no more in the way of instructions than "make it searchable on the Web."

Recently, I was presented with a user name and password by a publisher who wanted me to review his site. I went to the site, clicked "login," and was presented with a lengthy new user sign-up form. Not seeing any other options, I filled out the form, included the supplied user name and password. The system immediately rejected them, telling me they were "already in use." After much experimentation, I determined that while the site requires passwords, validates them and tracks them, there was no way to use them more than once. Returning users had to re-register from scratch and think up new user names and passwords every time!

Another site I reviewed gave new meaning to the term "keyword searching." I dutifully typed my keyword into the search box, and the system took me immediately to a document that was over 100 pages long, and left me there to, well ... search for my keyword ... as no highlighting was supplied.

In another remarkably crude search application, I entered a search term and got back 14,000 results. I was impressed until I realized the the site was returning unfiltered Google search results. Yes, this publisher was effectively offering paid access to someone else's free product.

This isn't a failure of programmers. It is a failure of management that left programmers to develop in a vacuum and let buggy products go live. To their credit, these publishers were aghast when I alerted them to these problems, and moved quickly to correct them. But what was most troubling to me was that in none of these cases had the users of these Web sites voiced any complaints. Why? Probably because they weren't in fact using them. And that's my point. If publishers don't take their own sites seriously enough to access them at least occasionally, they should not be surprised when their customers don't either. The act of simply "being on the Web" is not an achievement. In fact, a poorly-executed Web site is seriously detrimental to your business.

The New Year is traditionally a time for predictions, so I will dutifully take my turn. I am only offering one, but it's a big one: in 2006, we will begin to see a meaningful shift in the current GYM (Google, Yahoo, Microsoft) search hegemony.

Am I predicting the implosion or demise of one or more of the big three search engines? Not at all. Instead, I am predicting that the search business will begin to evolve in ways that don't play well to the strengths of the big general search engines and that we will see concrete evidence of this by the end of 2006. I also believe that we're starting to see the big three getting distracted, likely to the detriment of their core search business. Yahoo is increasingly emphasizing content, and has previously de-emphasized search to chase other opportunities. Google, for all its technological prowess and business success, is rapidly enmeshing itself in deals that will distract it, if not alter its very essence (think AOL). And at this late date, Microsoft is still trying to figure out what it wants to be when it grows up. You can never count Microsoft out of the game, but they're clearly running out of time and options if they truly want to be a major player in search.

Perhaps more importantly though, users are raising their expectations with regard to search. It was a huge technological achievement to get so much of the Web under one index, and I never ceased to be amazed by the new search tricks and innovations that are being rolled out on a regular basis. Yet, I believe the game is changing, and what users -- particularly business users -- hunger for now is not the most answers, but the "right" answer. To get to the "right" answer demands sophisticated filtering and true domain expertise, something that simply doesn't fit within the one size fits all framework of the general search engines.

If I am right (and at worst I may be off on my timing), we are going to see two types of search rapidly getting both attention and market traction: vertical search and paid search.

Vertical search is not a new concept, but it's had difficulty gaining mainstream acceptance and momentum, both of which seem imminent. While definitions vary a bit, to me vertical search means presenting a deep, highly filtered and very intelligent entry point into a specific market. Vertical search demands true domain expertise, because you can't filter intelligently unless you truly understand the information needs of a market, and you are experienced and confident enough to vet information sources to present the best as opposed to the most. Vertical search is most likely going to be advertising supported, but don't rule out paid search.

Paid search sounds a bit crazy, with GYM offering so much for free. Of course, people also said cable television would never succeed with broadcast television offering so much for free. I believe the marketplace increasingly wants both deep and dependable sources of information. This is where many of the content aggregators are playing right now, but most are hanging their hats on having the biggest collection of stuff, as opposed to the most comprehensive collection of stuff in a single area.

So while there will continue to be a large role for GYM to meet general and unsophisticated search needs, the future is about specialty search providers leveraging expertise in specific areas to become the search engines of choice for specific audiences. So while we are celebrating New Year's messages of unity in other contexts, in 2006 the data publishing industry should be welcoming -- and profiting from -- fragmentation.