4 Ways to Beat Google at Search

Today a new search engine launched called Cuil. Because of the resumes of the company’s founders — two are former Google employees — comparisons with Google have been fast and furious. Take a sample of this mornings headlines covering the launch: Search site aims to rival Google (BBC), New search engine takes aim at Google (CNET), Cuil Launches — Can This Search Start-Up Really Best Google? (Search Engine Land), Cuil joins Google, others in search for better search (LA Times), etc.

Fair or not, the the bar has been set super high by the media. So far, Cuil hasn’t seemed to have impressed many people, myself included. Results seem less relevant for many searches than on Google, and the site has been experiencing loads of technical problems in the few hours since its launch. The latter is forgivable — many high profile startups trip on launch day — but the first isn’t a great sign. Worse than Google won’t beat Google. Just as good as Google won’t beat Google. Even a little bit better than Google isn’t going to beat Google.

In order to beat Google on technology, you have to introduce some vast improvements to get people to switch. When Powerset launched in May, I put the site through a quick head-to-head with Google. The results were that for most queries, Powerset didn’t do a markedly better job of producing better results than Google. “If Google remains ‘good enough,’ Powerset will have a hard time convincing people to switch,” I wrote. Or in other words, unless your technology is vastly superior to Google’s, most people won’t be bothered to switch.

Cuil’s technology gimmick is the size of its index: 121 billion pages as opposed to Google’s estimated 40 billion or so.

Cuil also touts a better ranking algorithm based on “content and relevance” rather than “superficial popularity metrics” — a direct shot at Google’s PageRank. But their major selling point, their big “wow” factor, is their huge search index. The problem, though, is that competing with Google isn’t just a matter of competing with them on technology anymore. Google has become so deeply ingrained in the psyche of web users that getting them to switch to a new search engine is like trying to get people to use some type of plastic building blocks that aren’t LEGO.

4 Ways to Beat Google

Google isn’t invincible, but to unseat them from the search throne you have to do more than create a search engine that’s just mildly better at producing relevant results.

1. Create Far Superior Technology

The first way to beat Google, and the path that most search startups gunning for the big dog take, is to beat them on technology. But as I said earlier, to do that you have to be vastly better. I mean vastly superior. Cuil hasn’t impressed so far, and neither did Powerset. Yahoo! is trying to head down this road as well by opening up their search results pages to outside developers in the hope that the result will look better than Google’s offering.

However, for most queries, for most people, Google works well enough. It is rare that I can’t find what I need on Google — sometimes it takes a few refinements of my query, but generally Google comes through in the end. The same could be said for Live or Yahoo!, but Google still manages to get about 75% of my searches. Is their search tech that much better than Yahoo!’s or Microsoft’s? Nope. But it used to be, and in the time that it was I got used to relying on it, so now it is up to the other guys to come up with something better enough to make me want to switch. It hasn’t happened yet.

2. Disrupt Their Ad Market Stranglehold

“Google doesn’t make money on search – it makes money on advertising,” said ReadWriteWeb’s Alex Iskold a few weeks ago. And that’s an important point to consider because a second way to dethrone Google is to hurt it in the pocket book.

My former colleague Emre Sokullu suggested a way for Microsoft to do that in February. Sokullu’s plan called for Microsoft to split up their ad business into three parts: inventory, placement, and parameters, and then allow any third party to sell ads or services on any part of that open ad supply chain. The idea was to create a buying and selling environment that would be more open and attractive than what Google offers, and allow smart third party developers to come up with new and more effective ways to serve and target ads (in the same way that Yahoo! is hoping third party developers will come up with better ways to serve search results). That’s the sort of plan that is just crazy enough to work, and if it did it would have a serious impact on Google’s business.

3. Get to New Markets Before They Do

There will be 900 million mobile search users by 2011 accounting for $2.3 billion in ad revenue. Mobile web usage is exploding and beating Google to mobile search would be a huge win for anyone looking to unseat the kind of the search market.

Unfortunately for would-be challengers Google already dominates mobile search as it does web search. However, less than half of mobile searches rate their Google experience an 8-10 on a scale out of 10 according to Nielsen Mobile. So there is a lot of room for technology improvements in mobile search, more so than in web search.

4. Be Mediocre, But Be Persuasive

The final way to beat Google is to just pull in more eyeballs. This is where beating Google becomes less of a technology issue and more of a PR issue — it’s also the least likely path to beating Google. Microsoft, even though they’re in a distant third place, may be the best suited to attack Google in this manner.

As we noted last week, Live search is coming to Facebook this fall, which is a way for Microsoft to get in front of 80 million users with the flick of a switch. Even though Microsoft’s search technology is just “good enough” — or not good enough to get people to switch based solely on tech — if they can get their search box in front of enough eyeballs they could start pulling users away from Google. Unlikely, but plausible.

Free book: Jump Start HTML5 Basics

Grab a free copy of one our latest ebooks! Packed with hints and tips on HTML5's most powerful new features.

  • http://www.dangrossman.info Dan Grossman

    Cuil is strange. If you search for “web stats”, it puts W3Counter.com’s counter images next to the search results of 5 competing services.

  • Jonathan

    I don’t even know what to look at on the search results page. Way too much content preview without actual substance, and it seems to rank directory sites more highly than the site they are linking to in many cases.

    It failed to impress me and I gave it far more chances than the average searcher will. A lot of buzz for nothing if you ask me.

  • Tim

    Yeah, the images next to the links are scrambled, and the search seems temperamental and sometimes broken…

    I’m guessing one of their financers started pushing for a visible prescence, as this looks like a product that needed another few months behind closed doors before the hype-machine started working it’s magic.

    As it is, I’ve been, I’ve seen, and I probably won’t be back… at least for a while till the memory of the “meh” feeling dies away…

  • jhewitt

    Oddly enough, the cursor does not focus to the input field when the page loads(tested with Firefox). I would start improving there.

  • http://www.dangrossman.info Dan Grossman

    It’s certainly getting users from somewhere, as it’s starting to generate search hits to W3Counter-tracked sites, and the searches aren’t for site names or other stuff that would indicate only testing by site owners.

  • enter_oblivion

    If you search ‘image‘ in Cuil, with safe search on, one of the first results is a for LaunchImages.com which

    “is a free image hosting provider dedicated to providing reliable free image hosting where adult content is allowed.”

  • Goalie35

    I agree with Tim.

    Cuil needs much more fine tuning. It almost seems as if they tested the site amongst the four of them, and then went straight to launch. No beta testing or anything. I’m all for more competition in the web search market however I wasn’t able to locate many relevant websites when I tried my searches.

    I wish them the best of luck though!

  • cgustaveson

    I think it looks nice but the images are scrambled and odd (some of my sites have images not located on my site) and the results seem to be less targeted.

    I don’t even know how to read the results, which one do they say is the best left to right and down?

  • http://www.studio-gecko.com/ XLCowBoy

    I would place this as even pre-beta. Stage 1 even. I searched for the actual name of a site, and instead, got a 3rd party link to the site. Then I searched for a company product, and instead of cuil giving back the company’s official page, they gave me Wikipedia.

    Waaaaaaaaaaaaaaaaaaaaaaays off to say the least.

    What’s with the media nowadays? Even the “serious” ones are becoming nothing more than well-dressed hosts for sensationalist articles. What do they think their readers are? Tabloid’ers?

  • SlymG53

    How do you pronounce Cuil? That’s a big marketing mistake.

  • http://acquifolio.co.uk Online_Business

    i take it this only works for USA based sites currently? i also prefer my results in a single column, makes it easier to read in my opinion

  • weekbeforenext

    I just wish they would use an image from the web page in it’s description in the search results. It’s way weird that it just places random images. So much for image branding.

  • Stevie D

    I’m highly unimpressed by Cuil. I agree with Jonathan that it seems to favour directory sites – ie, the last place that anyone wants a search engine to send them. But not only does it favour them over proper pages – a lot of the time it doesn’t have any proper pages at all.

    Their problem is that they are focusing on the wrong issue.

    Google does not have a problem with not having enough pages indexed. Nor does Yahoo, Live, Ask or any of the other major search engines. We got past the days of search engines not having enough content years ago – a typical multi-word search on Google can return over a million results.

    The killer issue is now (and has been for several years) getting the most relevant results to the top of the list, especially given the enduring ineptitude of a huge proportion of web authors. As a general rule, I don’t care if Google gives me a dozen results or a million – what I want is for it to give me three to five spot-on pages in the top ten. By and large, it does this very well. Cuil doesn’t – it gives me absolute rubbish.

  • Desmond K

    No advanced search? No image search? And doing a search on ‘image’ lists Google and Yahoo! Image Search? The site looks fresh, but still needs some work.

  • fproof

    I just did some searches on Cuil that I did earlier today on Google, and I must say I found some very interesting and relevant results that didn’t appear on the first pages of Google. So I think I’ll give it a shot from time to time.

  • http://www.creativekitten.com/ ck88

    I can’t see any challenge whatsoever to Google in it’s current state. Very poor search results.

  • Anonymous

    Wheres the regional options??? It appears I can only search USA – or have I missed something?

  • tonixx

    What do you think about http://search.io ?