Searching for Quality
I'm in research mode for a client project right now, and that's all it takes to re-introduce me to the sorry state of site search - the capability a publisher provides for users to search just within the publisher's own site. The only conclusion I can draw from what I see is that publishers view site search as an expense to be minimized, and programmers view it as not worthy of their time and talent.
That's a big mistake.
Visitors to your site who use your site search capability are engaged users. They're investing time to probe deeper into your site because you've convinced them you are relevant to their interests and needs. These are prime visitors who you can monetize through advertising or a la carte sales of content. These are also visitors who'll likely come back again if you provide a rich and rewarding experience. Yet despite all this upside, site search is typically an afterthought. Here's my list of top offenses:
- Limiting yourself to low-cost or free site search software without regard to the kind of experience it delivers to the users of your site.
- Failing to integrate site search results into the site's style sheet and theme (usually because the designer is long gone before site search is implemented). Besides the jarring visual disconnect to those searching your site, your site screams "we don't care" which brings the value of your content into question.
- Providing only global site search capabilities. If your site contains different types of content, let users limit their searches. If I am on a newspaper site searching for a restaurant review, I don't want to wade through classified ads, news stories and obituaries in my search results.
- Not using content tags for searching. Most publishers are tagging all their content inside powerful content management systems to improve the front-end site experience, but I see little evidence that all these powerful CMS capabilities are being used to improve the back-end (i.e. site search) experience.
- No power searching capabilities. There's nothing more frustrating than reviewing results in order of relevance when date order is what you really need. Just a few simple options like this can radically upgrade the site search experience.
- No dates. A personal pet peeve of mine, there's nothing more frustrating than finding an absolutely on-target article through a site search and having absolutely no idea what month, year or even decade it was written, immediately rendering it worthless.
When you also consider that because many publishers hold their archives in database form meaning they are not typically visible to the big general search engines, site search gives publishers one shot at unlocking the value of this content. Instead, most are shooting themselves in the foot.
Fill 'er Up
A new launch by Picaphone, which aims to create the first international phone directory, caught my eye this week so I decided to check out www.picaphone.com.
After dozens of searches for all types of companies, big and small all yielded no results, I was pushed to an inexorable conclusion: this database nothing in it.
Back to the press release again and I noticed the statement, "The success of this ambitious project depends on the cooperation of web surfers all over the world." Yes, it's all about user-generated content, but with the remarkable goal of trying to collect every telephone directory listing in the world. Imagine how many listings would have to be contributed (and maintained) for this to become a site worthy enough for users to return to repeatedly.
I certainly have no beef with this company's audacious business objective. My point is that the window is rapidly closing for online data ventures that set up shop with a user interface and back-end database and then say to the world, "fill 'er up." Why? Very simply, the novelty factor is gone. That's why last year's Data Content Conference featured companies such as Snooth and BrownBook. Both these companies see user-generated content as integral to their success, but both started out supplying an initial dataset that delivered value while encouraging users to augment this information. With this approach, these companies deliver value to users immediately, rather than hoping magic will happen and users will do a credible (and rapid) job building out the database from scratch. I'd go so far as to argue that the bigger the scope of a user-generated database product, the more important to provide an initial dataset.
But what about companies like Jigsaw and Linked-In you may properly ask. These two very successful databases were built entirely from user-generated content. To this I would respond that these companies caught the wave of early excitement around user-generated content, so they had great timing going for them. I'd also argue that these databases, while they certainly became more valuable as they got bigger, were still able to deliver value to users while quite small. Think about it: both Jigsaw and Linked-In could deliver some value to users with 50,000 names; a database claiming to be a global telephone directory cannot.User-generated content certainly isn't dying; in fact where user-generated content augments a publisher's existing database it is very much alive and well. User-generated data products where the database starts completely empty are also lacking much promise.
Questions About Answers
A new enhancement from giant Canadian yellow pages publisher Yellow Pages Group caught my eye this week. It seemed to promise a live concierge service for users. You pose a question, such as "I'm looking for a good Chinese restaurant in Toronto," and you get a real-time response.
I surfed over for a look at the new feature, called "Yellow Pages Answers." Immediate disappointment. Here's still another publisher that simply grafted a social networking capability onto its database, with no apparent understanding of the fundamental dynamics of either social networking or yellow pages publishing.
First and foremost, social networking demands engagement, and that in turn usually means shared interests or some other compelling reason to participate and remain involved. Are there really people who find it cool, fun or interesting to be part of a community of yellow pages users? What shared interests do these millions of users who are doing quick information lookups have? Would you trust their recommendations?
Equally important, yellow pages is what I like to call a "point of purchase" medium. You only use the yellow pages when you want something, usually immediately. Submitting a question where you have to keep checking back for an answer, which might never come, is antithetical to the yellow pages raison d'être. The only people inclined to respond promptly are those with vested interests (owners of Chinese restaurants in Toronto, using my earlier example), or people with way, way too much time on their hands.
If Yellow Pages Group had launched a real concierge service that used it own staff to provide immediate answer, I'd be impressed. Certainly there would be lots of issues for an advertising-based medium to make recommendations, but at least there would be real user value. But encouraging your users to make vendor decisions based on recommendations from strangers comes with plenty of issues as well, not the least of which is how does this help my advertisers?
I hold up Yellow Pages Answers not to ridicule it but as an example that while social networking tools have their place and can be powerful and valuable, just because you can graft them onto your website doesn't always mean you should.
Labels: yellow pages answers, yellow pages group
No Reservations About OpenTable
OpenTable went IPO on Thursday, running up nearly 60% in its first day of trading (NASDAQ: OPEN). It was a much-needed "green shoots" moment for Wall Street, but it also speaks to the current favor in which subscription-based data products are held.
Does it seem odd that I am describing OpenTable, a nationwide online reservation system for restaurants, as a data product? We had a lot of discussion around that when we gave an InfoCommerce Model of Excellence award to Open Table in 2006. What swayed us in the end was its incredible workflow integration characteristics, coupled with all its data potential, much of it still unleashed.
Nobody could ask for tighter workflow integration. Restaurants that sign up for OpenTable have no choice but to make a total commitment to letting OpenTable manage all their reservations -- online and offline. You simply can't use it effectively otherwise. And once a restaurant has accustomed itself to OpenTable, enjoyed its convenience and seen the power of the online reservations pouring in through the OpenTable booking website, there's no turning back. Restaurants turn off OpenTable at their own peril!
But what about the data? Well, it starts with the fact that OpenTable is building a huge experiential database of dining activity. Registered users of the system even supply demographics, and OpenTable can easily overlay more. That means OpenTable knows with incredible precision, who eats where and when and how often. Think about an analytics and advisory business based on that data, much like MasterCard created MasterCard Advisors to leverage it analgous base of transaction data.
OpenTable is also well positioned to track no-show reservations, the bane of every restaurant. OpenTable could create a no-show score, much like a credit score. This would permits restaurants the option to either defensively overbook or perhaps demand a credit card in advance from such diners.
Open Table could help restaurants understand how they are doing versus their peers and competitors. It could help them learn how they are being discovered on the web. It could create customized promotion codes of landing page URLs that restaurants could use in their advertising to track response rates. It could tie into mapping data for site selection research. And the more it does, the smarter it gets, and the more valuable it becomes.
Intriguingly, Zagat Surveys was an early investor in Open Table, a relationship (presuming it still exists) that has been woefully underexploited. Tie the leading online reservation system to the leading rating system and an incredible number of opportunities emerge.
The key to OpenTable is that it successfully injected itself into an industry's order flow. Act One was to perfect the technology to make that possible. Act Two will be mining the rich data that results. From there, the future is OPEN.
(Since this may read as if I am playing stock tout this week, it seems right to mention that we have no business relationship or financial interest in any of the companies mentioned here.)
Labels: MasterCard Advisors, OpenTable, zagat
When Features Collide
Google's uses an annual press event it calls Searchology to trot out its new search features and other new offerings. Two new features caught my eye. Individually, they are useful and fascinating. In combination, they are potentially disruptive.
The first new offering, scheduled to go live imminently, is called Google Squared. It's an attempt to present search results to the user in spreadsheet format. You'll type in a standard free text query, for example, "baseball teams" and Google will return a spreadsheet where the rows consist of each team in baseball, and the columns are determined dynamically, such as location, league affiliation, etc. In short, Google is attempting to create structured data from unstructured web page content. Does it work? As you might suspect, it's far from perfect, but at the same time, it's a remarkable start.
Another new feature announced by Google is called "rich snippets" which involves putting more structured data in search results by drawing on web page metadata (in particular, pay attention to Google's support of a microformat called hCard). With rich snippets, which is being rolled out slowly, Google can now display star ratings for restaurants right in its search results, standardized address information for businesses, and disambiguated person data (initially courtesy of Linked-In) that will show company affiliation and job title right in the standard search results screen.
Okay, here's the disruptive angle to this. Fast forward a few years and it sure sounds that with more structured business and individual contact information available through Google, along with tools like Google Squared that allow output in spreadsheet format, it's entirely possible that users will routinely be able to easily create and export business mailing lists, organizational charts and much more. The more microformat data Google is able to access, the more high quality, structured output Google will be able to deliver to users and in a popular data interchange format to boot. Given Google's preferred price point for information (zero), there's a very real possibility of some disruptive change on the horizon.
Recommended response? Remember that basic contact information had been becoming commoditized for a number of years now. These Google offerings will only accelerate an existing trend. The future for data publishers is to continue to relentlessly push up the value chain delivering deeper, smarter information to customers in ways that integrate into their workflows and business processes.
2009 Model of Excellence Award Nominee: Boardroom Insiders
We're pleased to announced that Boardroom Insiders has been nominated for a 2009 Model of Excellence award for its "Boardroom Insiders" product.
Click here for our full MofE product profile.
Labels: google squared, hCard, Linked-In, microformat, rich snippets