There has been a lot of discussion lately about Google's forays into Vertical Search functionality, including Google Coop, OneBox and custom search engine support. Josh Kopelman of the Redeye VC blog had an interesting article on this topic recently: Google - The next vertical search engine? (I wrote about my take on it, in my previous article).
Vertical Search is one of the themes of this blog, and this particular topic - if/how much Google will affect the smaller, Web 2.0-based VSEs - is of paramount importance to the very existence of these engines. To get a more rounded discussion, I contacted some of the leaders in the vertical search space to get their perspectives. Without exception, each one replied with a measured, thoughtful response. The rest of this post presents and discusses those responses.
In my previous post, I mentioned SimplyHired - a VSE that provides specialized services and features catering to its particular niche users (job seekers). Gautam Godhwani, CEO of SimplyHired, responded to my Google/Vertical Search question, by highlighting SimplyHired's relentless focus on understanding and solving the user's real problem - in this case, that of finding the ideal job - something that keyword search, however sophisticated, cannot do. In his own words:
In general, I think that when a company like Google gets significant traction, there’s a tendency to think they can do everything well. As a generalist, Google will certainly try to ‘cherry-pick’ specialized features and implement them in its general framework. However, specialists (such as vertical search engines) that are successful continue to keep a specific user in mind, and provide deeper functionality than a generalist can possibly offer.
As any market matures, it creates more specialized offerings catered to specific audience. For example, in the social networking space, you do see dominant offerings such as MySpace. But look closer and you’ll see specialized offerings for each demographic (e.g. teens, college students, young adults, even baby boomers). I don’t believe there’s a secret sauce here. Specialists need to keep a relentless focus on the user, and of course, the segment needs to have a scope to actually create specialized offerings.
It turns out that ‘job search’ is much more than just typing a search query and getting results (though many of us wish that were the case). A successful job search includes researching companies, getting referrals, applying to companies, creating a resume, and various other tasks. As a result, there’s a deep scope in the job search area to create a specialized application, which has been the focus of Simply Hired since our inception.
Also notable is Mr. Godhwani's point about a segment needing to have the scope to provide specialized offerings; some categories/domains may be more defensible for VSEs than others. So things may not be black and white - it may turn out that Google will find varying success in different domains!
In a previous post, I referenced Retrevo - a vertical search engine for Consumer Electronics search. Vipin Jain, CEO of Retrevo also pointed to a strong focus on solving the user's underlying problem as a key differentiator for VSEs (interesting similarity to the SimplyHired response, although their domains are radically different). He used an analogy of sedans vs minivans - while a sedan is probably as close to a "universal car" as one can get, there are specific segments of the market that strongly prefer a minivan, and segments for sports cars, trucks, off-road vehicles and so on. [Personal note: Tell me about it - I have both a minivan and an old sports car; they serve very different applications!]. Mr. Jain says:
I wish all consumer needs (for specific applications) could be satisfied using keyword search and iterative search. If that were the case, we would all still be driving a black sedan, may be with some color options. People use sedan for everyday commute but they buy a mini-van when they have an extended family. They need more space (...) This requires changing the gut of an automobile. Adding a bigger trunk to a sedan doesn't cut it. A mini-van looks different, it feels different, it "is" different. People still own a sedan and they own a mini-van.
Segmentation is a natural part of any growing market whether it's cars, beer, toothpaste or information services. Information services that target specific industries and applications understand the gut of user needs: understand their environment, understand the relationship of information, understand the packaging and presentation of information (search is only one way to present it, other UI paradigms are needed as appropriate). All to serve a singular purpose: to understand a user's intent (clearly specified or derived from very high level abstract specification using various input schemes) and serve that intent in an intuitive way in the shortest possible time. Horizontal search engines can't do this without changing the gut and the presentation. And changing the gut and the presentation breaks the horizontal nature of these search engines. Yes, consumers need automobiles other than black sedans.
Regular readers of this blog will remember my earlier post about MetaMojo. Ashkan Karbasfrooshan, CEO of MojoSupreme (which owns MetaMojo) found this question on Google/Vertical Search so interesting that he wrote a long and thoughtful post about it on his own blog: Will Google Dominate in Vertical Search?. Here is an excerpt:
My approach to vertical search has been context. My third, upcoming book is called Context is King for a reason. MetaMojo.com’s approach to vertical search is radically different; we think that someone looking for [information] on Berlin during World War II wants one result while a couple looking for travel information on Berlin wants a different set of results, and we strive to offer the same user both. And, we’ll continually scour the Web to ensure that the results are very good.
Finally, Alex Iskold, CEO of AdaptiveBlue, (note: AdaptiveBlue is not a vertical search engine) has an older post on this topic on the Read/WriteWeb: Watch Out Google, Vertical Search Is Ramping Up!. In the article, he talks about the trade-offs:
There are always tradeoffs between generic and specialized applications, particularly in the area of information processing. The generic applications cover a wider spectrum, but specialized applications excel in their niche - because they have an intimate understanding of the semantics of information in their topics.
In the end, there will always be place for generic search - because it may not be economical to have a vertical search engine for every vertical. However, in major verticals, specialized search engines might take a big bite out of the generic search engines' market share - including Google's advertising pie.
What do all the responses have in common? Clearly, VSEs differentiate themselves with a strong focus on a particular user set and on solving domain-specific problems for that user set. Within their specialized domains, especially for more serious or complex problems (e.g. healthcare, real estate, finance, job search or even consumer electronics - with innumerable models from different manufacturers, each with varying levels of compatibility and format support), specialized engines should be able to continue to lead the larger, generalized search engines in utility. However, the larger search engines like Google, Yahoo and MSN lead in mindshare and are likely to continue to do so; power users may migrate to engines providing specialized functionality, but it's hard to imagine a general user keeping track of a dozen different sites, one for each vertical.
One possible direction is that search will become a "feature" of various applications and content sites, powered by individual vertical search engines, so that instead of heading to a search engine website, users can simply search from within the context of wherever they are, and it's up to the site or application developer to pick a high-value search engine. (This scenario was proposed to me by Tom Eng, founder of Healia - see my next post for his comments.)
Note 1: Alternative viewpoints
If someone would like to present a counter-argument from Google's perspective - for example, to make a case that Google can leverage its extensive infrastructure and huge user base to provide better services to end users, even within a vertical - please let me know and I will be happy to provide equal air-time (I mean, blog space!) for that perspective.
Note 2: Disclaimers
- In case readers are wondering: I do not get paid to write articles for this blog. [If that changes in future, I will mark the paid posts explicitly and clearly.] At this time, I also do not own stock in any company I've written about so far.
- The intent of this blog is not to be partisan, but to try to provide a balanced perspective. I certainly have my biases, which will be pretty evident to regular readers; but I'll be happy to provide space for anyone wanting to make counter-arguments. For example, I'm inclined to be sympathetic to the smaller, focused VSEs, since they can solve small, specialized problems very well. In a sense, VSEs represent a further extension of Google's "do one thing, but do it very well!" philosophy. My overall goal is to start and continue an ongoing discussion, one that hopefully contributes to making technology better for us all.
I found a related article by Dave Kellogg of the Mark Logic CEO Blog: "Give Me Back My Google". It appeared over this weekend as well - an interesting coincidence in timing.