In his recent post on the Redeye VC blog (of which I am a regular reader), the always interesting Josh Kopelman has formulated an interesting premise: can Google become the next Vertical Search Engine?
In particular, he focuses on Google's OneBox functionality as an attempt to provide vertical search features
What I've been most interested in, however, is Google's approach towards vertical search. While a lot of attention has been focused on the recently launched Google Co-op -- which allows Google users to create a search engine for a customized collection of content -- I think that not enough attention has been focused on their OneBox feature. I think Google OneBox is a pretty decent attempt to create vertical search functionality inside of Google web search.
Josh then goes on to give a boatload of examples of Google's specialized features in various vertical domains - travel, books, news, people, shopping, restaurants, and so on. If he's right (and judging by the number of successful companies he's started, he very often is), then the implication is that Google would not only become a VSE, but in fact (being Google) would become the mother of all VSEs, the vertical search engine for all of us.
In highlighting the increasing importance of the OneBox functionality, Josh is right on the money. As is usual with Google search products, the powerful capabilities of this function are masked by its apparent simplicity (the image of an iceberg comes to mind) - in one stroke, Google is able to offer context-driven, authoritative and relevant search results, features that so far have been the hallmark of Vertical search engines. There is no doubt that VSEs will continue to see Google looming ever-larger in their rear-view mirrors as it catches up on some of these features.
On the other hand, by concentrating on specific industries and verticals, Vertical search engines can utilize their deep domain knowledge to provide so much more - deeper features and services that a general-purpose search engine like Google or Yahoo would be hard-pressed to match, such as: domain-specific display services, additional data integrated with search results, parametric search, related services and so on. [I've described some of the VSE differentiators in an earlier post].
For example, SimplyHired, one of the leading vertical search engines in the job search category, provides a host of special features, such as: resume-posting services, additional datasets ("dog-friendly workplaces"), filters ("company size", "working mother friendly"), email alerts, salary research, integration with LinkedIn, job search forums, job suggestions (a la Netflix) and even SimplyFired (you'll have to check that one out for yourself!).
Josh also makes the same point in his post
I think that vertical search engines can still succeed if they offer additional application-specific features/navigation that differ from traditional search-box functionality (like Krugle) or if they focus on syndicating their results across other sites (like Biggerboat).
Overall, it's hard to imagine Google providing this level of service and integration in every vertical domain, or even in a large number of domains. On the other hand, Google has one trump card that most of these VSEs do not have at this point: a dominant mindshare among end users. Hmm - do I sense some "strategic acquisitions" coming up in this space?
For the long term, though, I think Vertical Search Engines are here to stay; those that provide specialized, useful features and services for their domain will continue to co-exist happily with the larger, more general search engines and potentially even continue to grow!
Will Vertical Search and Social Networks combine to challenge Google?
Publishers and advertising agencies have a very difficult challenge ahead as traditional “horizontal” media like newspapers, TV channels and magazines see their traditional demographics and advertising revenue streams fragmented by the increasing preference of consumers for online access and the huge presence of Google eroding their audiences and potential future revenues.
Perhaps they should remember the words of Sun Tsu, who once said “When the enemy is too strong to attack directly, then attack something he holds dear. Know that in all things he cannot be superior. Somewhere there is a gap in the armour, a weakness that can be attacked instead.” Google’s major strength – the clean search box and the ease of use, commoditised ad revenues, perhaps masks its principal weakness. As media content and advertising revenues fragment to serve thousands and thousands of “vertical” online communities based on lifestyle or profession, Google may suddenly seem standardised, commoditised and lacking a sense of unique community. Is Google becoming Wal-Mart, while vertical communities may prefer Harrods?
Whilst “horizontal” media companies are similar to supermarkets, specialist professional “vertical” publishers are very specific in serving niche communities with totally relevant content and requirements. However, the publisher’s principal operating difficulty in becoming adaptive to this asymmetric Web 2.0 opportunity is that most tend to run each of their print, exhibition and online titles/businesses as separate profit and loss items on their balance sheet. As a by-product the vast majority tend not to have a centralised IT infrastructure or the human IT skill sets to manage a large scale data centre or web spidering facility – the prerequisites needed to datamine and aggregate open source, user generated and blog content to create vertical slices of the Web that are relevant for their audiences. Publishers will also need to integrate this content into the online extensions of their print brands and thereby allowing advertisers the opportunity to target high value communities. In addition, the datamining, crawling and hosting to identify relevant open source content will also need to be a continual process due to the continual growth of user generated and open source content.
Convera have two very large data centres, an extensive web spidering capability and a web index. Convera are now partnering with a significant number of specialist B2B publishers to create a range of vertical websites for specific professional communities. The first example of this is Searchmedica.com with UBM.
In building the deep vertical search portals, the key is to reach into the specific professional community in a number of ways. First, you can combined the trade publisher's knowledge and contacts in the profession with community appeals that engage the specific audience in a way that general search cannot, and also by taking special care to use the taxonomies common to the targeted profession in organizing search results so that the user feels more at home and among peers. Building a good vertical engine can be costly and time consuming, and getting a critical mass of users to de-Google their search habits into more specialized engines is potentially a tough sell. However, in tests with focus groups from different professional communities to test these vertical search properties against Google, the results are hugely encouraging.
In building the beta test sites, the specialist publishers are providing Convera with "white lists" of data sources online and websites that would be most relevant to its readers so that the searches are restricted to reliable and trusted information. Publishers are also securing agreements with owners of key proprietary content not normally crawled by Google by leveraging some of its contacts and resources so that Convera can crawl and deliver some of their proprietary content. Another key consideration is getting the user community engaged in the process as co-developers. No matter how bad the results at Google or Yahoo may be for a given professional segment, the interface is familiar and the destination is always at hand. Getting users to think of a specialized brand as the go-to place for business information is the challenge.
A number of publishers are actively assessing the potential of adding social networking to the mix in order to get professionals interacting with each other and adding weekly podcasts by industry experts on issues affecting the community – these additional services will create more community loyalty and also additional advertising and sponsorship opportunities.
The publishers can also use their print titles to drive the audience to the new online areas and this will also assist the transition of their high value print ad revenues to online. Publishers also have exhibitions, seminars, events and email newsletters to assist this transition – and recent research suggests that professional communities will actively attend seminars and events to meet peers and other members of their community. The theory goes that once you get some professionals involved then the viral mechanism or behavioural “Hive Mind” also kicks in and professional workers start referring to the vertical portal as a community source. It is also allows advertisers and public relations organisations access to a clearly defined, affluent, influential and stable audience.
Google does not allow you to have a beer with a potential business partner - it doesn't have that sense of community. But Google is fighting back – the recent launch of Google Custom Search and acquisition of teenage social network sites indicates they are aware of their weakness – but specialist publishers see this as a Trojan Horse. Social networks for teenagers are highly transient and target a demographic that is volatile, unpredictable and has a low level of disposable income – whereas a social network alongside a vertical search service for 22,000 bio-chemists, 55,000 UK GP’s, 55,000 insurance risk assessors or 120,000 US psychiatrists is stable, affluent and attractive for advertisers.
Posted by: Andy Black | March 19, 2007 at 05:17 AM