Tim Houghton Moving Up: Vertical Search Proliferates
Jinfo Blog

Thursday, 1st March 2007

By Tim Houghton


Vertical Search is one of the 'hot' areas of the last couple of years. Not that it has dominated Google's overall most popular queries. Britney and Paris see to that. But quietly amongst thoughtful commentators like Om Malik, Danny Sullivan of Search Engine Watch and the financiers of Silicon Valley, the area has attracted expert attention.


Tim HoughtonVertical Search is one of the 'hot' areas of the last couple of years. Not that it has dominated Google's overall most popular queries. Britney and Paris see to that. But quietly amongst thoughtful commentators like Om Malik <http://www.gigaom.com/>, Danny Sullivan of Search Engine Watch and the financiers of Silicon Valley, the area has attracted expert attention.

But what exactly constitutes vertical search, and how has it been executed on a consumer and professional level? Below are a few thoughts on what is driving the growth of vertical search and whether it is with this rather than general search that the future lies, especially as far as those who use the Web for work are concerned. These are primarily sources that utilise online content and present this via websites. There are obviously many other potential types of vertical search.

Defining vertical

You'll be relieved to know that defining vertical search is considerably more straightforward than defining other hot topics of the moment such as Web 2.0. As Wikipedia puts it, 'Myriad specialized search engines are emerging to address the particular information needs of niche audiences and professions.' Om Malik defines it slightly differently: 'It is a specialized search engine that mines data for one narrow niche of the market place.'

Another aid to understanding vertical search is to understand what it is not. Google's stated mission 'is to organize the world's information and make it universally accessible and useful'. And clearly rivals like Yahoo!, MSN, Google and maybe even open-source competitors like Nutch have the same universalist aim. Clearly no search engine gets even close to achieving this objective today; they don't even index the entire Web, and, as Google discovered when it started scanning books still within copyright, there are formidable legal as well as technical obstacles to doing so.

So a vertical search engine is trying to solve a different, more specific problem than a generalist one, focusing on the needs of a specific market segment, user group or alternatively a highly specific dataset. Is this a viable strategy or will these minnows be crushed by the larger players? Is search tending to atomise, to become complex and fragmented, or is it set to be dominated by a few giants?

Consumer offerings

To help understand this, here are a few examples from the consumer space, before examining the professional sphere. Technorati http://www.technorati.com/ is now almost too big to be considered a 'vertical' search engine at all. But it certainly started as a niche offering; it went from being a somewhat clunky blog search engine to a highly polished offering that specialises in the indexing of user- generated content. In the process it showed that even in quite a broad content area Google can be challenged.

Technorati is 'vertical' in the sense of being focused on one data type - blogs - but its target audience is broad. Likewise job search engines like Indeed and SimplyHired focus on a smaller, though still vast dataset, job postings. But their audience is also restricted, though not by demographic or geography, but by intent. Try a search in either of these engines versus, say, Yahoo!, and it is hard to argue that they're not yielding a higher quality set of results. And one can drill down further. If you're looking for a job in the information industry then one can argue that Jinfo is a high-quality vertical search engine. This also raises the issue of where a focused portal ends and a vertical search engine begins.

Take a quick glance at other successful consumer vertical search engines and a clear pattern starts to emerge. Areas like travel (Kayak and Farechase, now owned by Yahoo!) and property (FindaProperty, PropertyFinder) are well-established players, and automotive looks like the next area for growth. In the US one of the leading online car sales sites recently announced plans to open an automotive search site, MyRide.com.

Clearly, in the consumer vertical search area, searches that involve complex informational requirements, high levels of engagement and large amounts of data work well for vertical search providers. It is worthwhile for the consumer to go to the trouble to seek out such a service. And notice too that these sectors are among the most buoyant for online advertising and not coincidentally are characterised by high levels of classified advertising in the traditional print world. Site creators realise they can deliver a more focused and hence higher-value audience to advertisers than general search engines.

Professional use

But what about the professionals? As 'super searcher' Mary Ellen Bates has long argued, it is well worth looking beyond Google. In the professional sphere one hopes and expects that those using the Web for work will be willing to invest the time both in identifying and learning how to use more specialised services.

Of course vertical search for professionals did not arrive in the mid 1990s with the Internet. Databases of carefully filtered, validated and trusted content have been around for far longer - LexisNexis and Factiva being just two obvious examples. But vertical search that specifically utilises Web-based data is a newer concept.

SearchMedica, a newly launched search service focused on the general practitioner sector, illustrates the demand for sector specific information for professionals. It's interesting to note that the service is not built from scratch using its own technology but instead combines an indexing partner, Convera, with domain specific experience. And there are many others, eg High Beam for market research or Alacra, a meta-searcher to over 100 structured databases.

One of the most intriguing new services if you're a buyer of information for investment professionals is Monitor110. Its aim is to be a complement to a Reuters or Bloomberg terminal for traders. How? By providing actionable investment insights from user-generated content. Roger Ehrenburg explains on his blog Information Arbitrage that by focusing on one sector and using experts in that sector you can predict with a high degree of certainty what users are looking for and how they would like information presented.

But take that one step further. Why stop at a sector when you can focus a search on an individual business? That's what David Seuss at Northern Light has done over the last few years as the firm has moved from providing a fairly general search engine to focusing on providing what I would call ‘enterprise specific vertical search’. When I spoke to him he enthusiastically explained how the firm provides a bespoke research portal to clients, covering both internal and external sources with user-level access permissions and so forth. As he puts it, 'The future is bright for complexity.'

Upward growth

In some ways the growth of vertical search alongside horizontal or generalist search is utterly unsurprising. As the number of Web users grows to 1 billion plus, their experience and sophistication grows, as does the amount of time they spend online. It is hardly surprising that tools proliferate. As Clay Shirky points out in a recent Wired News article <http://www.wired.com/wired/archive/14.11/meganiche.html> a 0.1% market share of Web users is 1 million unique users. And if these are high-value professionals, that is potentially a highly profitable market.

In general I think any analysis that boils down to a simple 'either/or' dialectic is likely to be wrong. In the case of vertical search I think this is undoubtedly the case. As the vast scale and complexity of Web information grows the big players like Google, Yahoo!, MSN and Ask will continue to have tremendous scale economies when it comes to indexing and serving general queries from the Web. But for exactly the same reason I think those demanding intelligence rather than simple information will use other tools.

So will the major search engines be reduced to mere fact checkers? Not at all. I think they will develop vertical offerings in the larger consumer niches, as indeed they are already doing. But skills like the ability to support complex structured searching, taxonomies, access to databases and complex data visualisation tools all have a healthy future for those services targeted at information professionals. Interesting that Silicon Valley is backing ideas that information professionals have argued in favour of for years.

Related FreePint links:

Related links:

« Blog