Meditations on programming, startups, and technology
New Relic

A tale of two search engines

Remember when Altavista seemed good enough? Then along came Google and seemingly overnight everything changed. We didn’t even know that it was possible to receive such good link suggestions from a search engine. Yet there, right before our very eyes, it happened.

These days, highly popular search engines are worth billions of dollars, mainly thanks to the massive advertisement businesses that can be built on top of them. The incentive to get a slice of that huge pie is clearly there. But can we do better than Google? And will such innovation necessarily arrive from the research labs of giants like Apple or Microsoft?

The answer to the first question is obviously yes. There is always room for improvement. The latter question may appear equally obvious as well, but let’s take the tale of two different search engines into consideration.

One of them was started by xooglers (ex-Google employees) with plenty of experience in the field of search engines; it had a team dedicated to its development and could afford to have a VP of communications. They received plenty of funding ($33M) to get the ball rolling, and garnered a fair bit of press coverage when they first launched.

A very different search engine, however, was started by just one person. It was bootstrapped (no external investors), came wrapped in a silly name, and virtually no one paid attention to its launch.

The first is Cuil (pronounced ‘cool’), which as many know, has become something of a running joke online. It’s the perfect example of how not to create a startup and of everything that could possibly go wrong with an ambitious software project.

From day one the results were so incomplete and irrelevant, that one has to wonder if the $33M they received was spent on developing the technology needed to clone Samuel Becket and place him in charge as the chief architect of the project. The absurdity of Cuil’s search results even led to the development of a highly entertaining Cuil Theory.

A few days ago the Cuil team launched an automated Wikipedia of sort, called cpedia. The end results were so terrible, that they will serve as an eternal cautionary tale against the indiscriminate use of Markov chains.

Unless Cuil/Cpedia is a practical joke aimed at the tech community, they may as well shut them both down. At this point they really are just wasting their time (and ours).

The second search engine, the one with a rather wacky name, was created by a single person (Gabriel Weinberg) and is called Duck Duck Go (a play on the name of that old childhood favorite, ‘Duck, duck, goose’). However, much to the surprise of many, there is real innovation going on here (particularly presentation wise).

Duck Duck Go

Duck Duck Go’s zero click information is very useful – as is dividing the results by topic (for example, the word “ruby” can have different meanings in different contexts). Presenting all the search results on a single page (via AJAX) was a smart and handy feature as well. With this search engine your privacy is respected, particularly since your IP is not even logged.

This site is still as niche as it gets, but it’s reaching a tipping point amongst the ultra-geeks — not to mention that more and more people (myself included) are adding Duck Duck Go as their default search engine within their browsers. Is it better than Google? No, not always. It depends on the type of query. Sometimes it’s better, sometimes it’s worse, but it’s usually quite usable and is a concrete attempt to innovate the search engine realm.

Duck Duck Go serves as a poignant reminder to the software world that David can still strike Goliath.

No related posts.


If you enjoyed this post, then make sure you subscribe to my Newsletter and/or Feed.

receive my posts by email

7 Responses to “A tale of two search engines”

  1. Thank you so much for the nice write-up! And please let me know if you have more feedback.

  2. matt says:

    OMG, the cuil wiki results are bad! I hadn’t seen them before. The ruby language page you linked to was bad, but readable — a few searches yielded more examples that fail even the readability test.

    Clearly there’s not enough “I” in Cuil’s “AI.”

  3. Revital says:

    A very interesting comparison. I think Duckduckgo is a very good engine, and certainly a refreshing breeze in the search engine arena.

  4. Congratulations Gabriel. As a fellow Philly area entrepreneur, I’m inspired by your success!

  5. Stuart says:

    I had to laugh at Cuil’s own page on cpedia:

    http://www.cpedia.com/search?q=Cuil

    “Hell no..!! Cuil is a search engine that has launched on this (28th 2008). “But according to reviews on CNN, cuil fails to deliver a good service better than Google.”

  6. Aaron says:

    I’ve been rockin’ DuckDuckGo for a while now and love it. Gabe is a trusted colleague and brilliant entrepreneur. He could conquer a whole team of GooglyGiants.

  7. Robert says:

    Hi Antonio,

    I see that your blog is occupying second place in the search result group “Ruby (programming language)”. Pure coincidence of course. =D

    Seriously though, I had never even heard about Duck Duck Go Go. I already love it though. The grouping keeps the search results nicely on-topic.
    I wish that I could search within results though.

Leave a Reply

I sincerely welcome and appreciate your comments, whether in agreement or dissenting with my article. However, trolling will not be tolerated. Comments are automatically closed 15 days after the publication of each article.

Copyright © 2005-2012 Antonio Cangiano. All rights reserved.