Vertical Search Presentation from the EPublishing Innovation Form

I had the pleasure of attending and speaking at the EPublishing Innovation Forum in London this week, chaired by David Worlock, chief research fellow, at Outsell, Inc.

I participated in two sessions, the closing panel with Hugo Drayton of Phorm (a controversial behavioral ad network that collects data at the ISP level), Vin Crosbie of Digital Deliverance (a management consultancy to the media industry, particularly newspapers), and Leonard Brody of NowPublic (a citizen journalism news website fueled by crowdsourcing).

My primary participation, however, was in a session entitled Risks and Opportunities in Vertical Search, where I spoke with Iain Fletcher from Convera, a company that has wholly, boldly, and some might argue, insanely (given that they sold off 93% of their revenue stream), repositioned itself as a vertical search platform provider. (As I always say, I give Convera an A+ for commitment, if not a similar grade for strategy. The concept is interesting, by the way, but the question is can it support a viable ongoing business?)

My main messages were:

  • The risk in vertical search is kidding yourself into thinking that you can make money with a site that’s nothing more than you can throw together in a half-day using Google Custom Search Engine.
  • The opportunity is in adding value. I focused a lot on audience intimacy and value because that’s where I believe publishers can build safe, profitable, sustainable markets in the face of horizontal forces such as Google.
  • I said that I disliked the term “vertical search” because I think it sets too low a bar. To most people, I think vertical search means a domain-specific crawl and returning results list of links to documents. That’s not enough.
  • I echoed the thoughts of Andrew Richardson from Wolters Kluwer Health who spoke the day prior and, among other things, argued that contextual design is an appropriate methodology for building online information services.
  • I briefly argued in favor an XML platform approach to delivering such products and services. (Because you can put all your content in one place and then XQuery to quick build products which should be viewed as experiments — fail early and often.)

In summary, my vertical search talk said: don’t think about vertical search; think instead about content applications. Don’t build search sites that are low in value-add. Build applications that help users accomplish tasks.

Here, courtesy of Slideshare, are my presentation slides.

One response to “Vertical Search Presentation from the EPublishing Innovation Form

  1. Great post on vertical search! Convera as you mentioned is putting all its eggs in the vertical search basket. However they are primarily focused on working with publishers to search enable their vertical content. There is a long tail market for vertical search that spans across publishers or content domains where a a number of smaller players can put together vertical using lucene/nutch or content search applications based on them like SearchBlox. And it is not just about putting a search together but providing value add around these vertical sites. Now with the availability of computing/application resources by the hour on the cloud computing platform, the barrier to creating custom vertical apps requires little investment.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.