Archive for the ‘Google’ Category


Bessed, Mahalo and Human-Powered Search

July 18, 2007

I’ve been asked about it several times, but have been tardy in writing about Mahalo, the new human-powered search engine that Jason Calacanis is spearheading. I did get a chance to talk about it a bit with the New York Times‘ Randall Stross, who wrote this piece on Bessed, Mahalo and other search competitors a few weeks back.

I’m grateful to Stross for including Bessed in his piece, as my initial fear about Mahalo was that people would think that Calacanis had thought this up all on his own, and had thought of it first, when in fact Bessed was launched in October of 2006, long before Mahalo. I was afraid people would think that we were the copycats.

To his credit, Stross did his homework. He realized that Bessed had launched this concept of a “human-powered search engine” before Calacanis came out beating his chest and talking up the VC money he has backing him up.

I’m not upset about Mahalo launching almost a carbon copy of what Bessed is doing—or, as a friend e-mailed to me, “Dude, they stole your idea!” (Although it was a litte disheartening to see them tout themselves as the “first human-powered search engine.”) That’s the nature of competition. Frankly, I’m jealous of the money Calacanis has behind Mahalo. It will be interesting to see what it gets them.

However, there are some differences between Mahalo’s game plan and that of Bessed, and I think those differences are what will ultimately doom Mahalo, or at the very least force it to change course from it’s currently-stated plan. I’m also afraid Mahalo might kill the idea that human-powered search can work, because its current offering doesn’t offer a ton of value. And if that happens, it could hurt Bessed over the long run. So, while I would not be unhappy to see Mahalo fail, how it fails matters to me 🙂

First, here’s what is good about Mahalo. (Generally it’s the same as what I think is good about Bessed.) Mahalo is having human editors find results, which is eliminating spam from its results. The site looks attractive. It’s allowing visitors to suggest new sites to add. And I think it offers good results for the topics it’s covering.

But Mahalo makes one big mistake. It is attempting to create results for only the most searched-for terms. The problem is, most people are perfectly happy with Google results for the more common searches. They aren’t looking for an alternative. Where Google and other engines often fall flat and and are susceptible to spam is in the “long tail” of searches—searches for specific people, products, facts, etc. These are the searches in which searchers come away dissatisfed and are open to an alternative that can solve their problem and save them time.

I don’t know if any human-powered effort can adequately cover the millions of potential searches that take place each day, but by simply ignoring them Mahalo has no compelling reason to exist. It does not solve a searcher’s problem, so beyond what Calacanis can drum up traffic-wise based on his own personal celebrity, it will fall flat.

Our goal with Bessed is to fill the holes in the long tail, sifting out the junk on those specific searches that so often are maddening—when you find one site selling the same thing on four different domains or you are lured to a site on false pretenses because the site has pasted your keyword (and a hundred others) on a page that is completely irrelevant. Those searches drive you crazy, and Google’s algorithm, which puts so much stock in the links between sites, has a hard time sifting the junk because there are so few links between sites in the long tail, thus making it hard to give any of the pages credibility over others. This is where the humans can and should be; this is where we can make a difference.

This doesn’t mean Bessed will ignore the “short tail,” but it means we know that we can create more value in attacking searches that robots have not yet mastered. If I had Calacanis’ money, this is where I would be spending it. Maybe he’d like to give it to us?

e-mail me:

Adam Jusko is founder and CEO of Bessed, a Web site promising “search without spam”, thanks to human-edited search results and ongoing visitor feedback. Do a search, offer your comments, submit your site–help create the “bessed” search site in the world.


Reporting the News in a New Way

October 24, 2006

My biggest concern in creating Bessed (other than money, time and resources) has been to differentiate it from what other search services offer.

We can not compete with the major search engines on their turf. We’ll never be able to supply the phone number of your local pizza place in every city of the United States. There is an inherent limitation to using human editors.

We can easily become the best Web directory, but that’s a short hurdle. It will take time, but I have no doubt Bessed will outshine the Yahoo Directory and the Open Directory over the long haul.

But that’s not enough. To be more than a small, marginally profitable entity, we have to offer something no one else is offering. To do so, I’ve looked at Wikipedia for inspiration. Bessed is very different from Wikipedia, but we can learn from it in building our site.

People love Wikipedia because it’s one of the few places online you can go to get a complete overview on a topic. Want to know about the magician David Copperfield? (Just pulling a name out of the air here, couldn’t say where it came from.) Wikipedia will tell you where David Copperfield was born, his life history and a few other interesting tidbits. Sure, the nature of Wikipedia means that someone could come in and write that Copperfield sleeps with 75 snakes each night, but in general you’re going to get pretty accurate information from the site. And no one other site is as thorough on so many topics.

What does Bessed do to similarly differentiate itself? We first must understand that to be a different kind of search service it won’t be enough to just offer a bunch of links. Instead, we have to paint a picture on each search results page. Someone should be able to come in and within the first dozen links have a great overview of a topic. Then, if they want to delve further, they can use the links further down the page. In the end, they should feel like they got everything they needed in the shortest amount of time possible.

No search engine or directory does an adequate job of this. Search engines use an algorithm that often gives you the almost identical content from 10 different competing sites, forcing you to keep looking and looking to get a clear picture of a topic you’re researching. Web directories give you a bunch of static links listed alphabetically and tell you to do with them what you will. Search engines aren’t ideal, directories aren’t even close.

Bessed has the opportunity to paint that picture. As an example, see our World Series 2006 results. We start with the official site of the World Series from Major Leage Baseball, but then the links are sort of a backward-looking view at the Series as it progresses. Right now that means that you’re getting Game 2 highlight links, including multiple links from news organizations and blogs about the controversy over whether Detroit Tigers pitcher Kenny Rogers was using an illegal substance to get a leg up on the Cardinals. Further down are select links on Game 1, further down from there are links to articles that previewed the Series. Interspersed among those links are links to commentary on the Series overall. By the time the Series is over, a visitor will be able to come in and quickly find just about everything he or she would want to know about this World Series, from the big picture to the game-by-game highlights to the opinions of both journalists and fans who followed the event.

We’re not creating content, but by artfully putting the story together through the use of timely, relevant links, it becomes a comprehensive look at the many aspects of this World Series.

Our goal is to do this as much as possible, to become the go-to site for anyone who wants a well-crafted overview of an event, person, subject, etc. Some topics will lend themselves to this better than others. The World Series has more depth than the subject of air compressors, unless you’re really, really into air compressors.

If we can become the go-to site for the best links on topics people are looking to really dive into, we will also get the less deep searches as well, and that’s what will ultimately drive our success.

I mentioned to someone not long ago that what Bessed is doing is what a smart newspaper would do. A smart newspaper would continue to report stories, but backfill those stories with links to other newspapers, blogs, sites of interest that provide an overview of the subject being reported on—each story offers the reader a chance to get the big picture. No one does this because they don’t want to give up the eyeballs. (To be fair, some newspapers have begun to include blog links alongside stories, but it’s an automated, hit-or-miss affair.)

We’re certainly not going into the newspaper business, but we’re going to fill that gap where search, news, blogging and wikis intersect—or where they would intersect if anyone was willing to put the work in to make it happen.


Is Google Custom Search a Revolution or Just Keeping Up with the Yahoos?

October 24, 2006

As part of its Google Co-op offering, Google is now offering you the ability to “roll your own” search engine. Basically this means you can create a custom search engine for your site that can limit the search to certain URLs, or give certain URLs priority in the results. This can be used simply as a site search to allow people to search your site, or you can include many multiple sites in your engine that are authority sites for whatever niche you might be interested in. For example, you could create your own Beatles search engine and give priority to sites you know do a good job of focusing on Beatles content. And, Google’s offering will add Adsense ads to the search engine, giving Webmasters a share of revenue if site visitors click on the ads.

It’s a nice offering, but I imagine it will have limited adoption. Why? Because it takes work to create a custom search engine in terms of deciding which URLs are more important than others, and unless you really see this search engine as providing a service to your visitors, you probably won’t bother. Yes, there is the revenue share aspect that will drive some people to it, but God knows Google ads are already everywhere. Webmasters don’t really need a new custom search engine as a way to increase their Adsense income–the result is more likely to be a shift in terms of where the revenue comes from than an increase in revenue overall.

The most obvious application is for Webmasters to use Google’s search interface to offer visitors the opportunity to search their own sites, something Google’s already been offering for a long time.

I don’t know if this is really seen as a big idea inside Google or simply a way to keep up with Yahoo’s Search Builder and similar offerings from companies such as Rollyo.

The fact is that in most cases when you do a search, you don’t want it limited to just a few sites. Although it may be true that you’d like to see sites come out on top that you consider most trustworthy, you don’t want to shun sites that are little-known but happen to cover small niches really well. And it’s not always readily apparent where a good search result is going to come from. A site that spends a lot of time talking about hockey might suddenly write a very informative page or blog post on the best places to visit in Toronto. If your personal search engine excludes certain sites or always makes others more important, you might miss out on the good stuff from unlikely sources. Some will be willing to pay that price, but most will just continue to let the search engines sort it out, even if at times the results they get are frustrating.

TechCrunch and GigaOM both covered this, and neither seemed to think it was overly earth-shaking.

Via Brewed Fresh Daily, here’s a Google Custom Search just created for Cleveland, Ohio information.