- The most used content type among knowledge workers for business purposes has switched to press releases.
- People are spending more time (not less) searching for information online.
- Internet searches fail 30% of the time. This high failure rate suggests that further search engine specialization and vertically-oriented portals might make information easier to find.
The first point is a big change in the B2B/trade segment of the publishing industry.
In the old days, companies wrote press releases but no one saw them directly. You used a wire service to send them to journalists. Sometimes, journalists picked your story up off the wire and wrote about it. But usually you had to “sell” your story to them in advance by calling them (before you put the story on the wire), trying to pique their interest.
Once hooked, you’d provide an executive interview, 2-3 customer references, and an analyst reference or two. The journalist would take your press release, add a quote from 1-2 of the interviews, write a short story, and publish it.
Sometimes the journalists would add a lot of value in this process. Often, particularly in the weaker books, they did not. What’s happening now is that the latter group is being disintermediated by corporations.
The Internet played a huge role in this transformation by enabling the direct distribution of press releases.
I think smart companies have played a role, too. At my last job, when we identified this trend we started to create two types of press releases: (1) standard releases and (2) story releases. A story release was written to emulate a trade press article, right down the cute headline (e.g., Ben & Jerry’s Scoops The Competition With BusinessObjects).
Let me talk briefly about the last point cited above: “this high failure rate suggests that further search engine specialization and vertically-oriented portals might make information easier to find.”
I concur. In fact, it’s what many of our customers are doing with MarkLogic. Instead of trying to “make search better” they are building content applications that contextualize content (i.e., put it in the context of a task) and leverage a mix of proprietary, open, and licensed content.
You might think of these as vertical portals. But indeed I’d argue that (the good ones at least) are really content applications. It’s not about limiting the crawl to one domain and providing a regular search engine on top. It’s about figuring out what people are trying to do, putting their queries in that context, and returning content directly, not just links to it.