Duck Duck Go is a search engine created by “serial entrepreneur” Gabriel Weinberg. It’s at http://duckduckgo.com or http://dukgo.com.

Its search results are drawn from a combination of sources, including its own crawler, Bing and crowdsourced sites like Wikipedia.

I started using it May 11. Here are a few notes on my initial impressions in no particular order.

Pros

• Duck Duck Go is fast. Very fast. As the Web editor of several news sites, much of my searching is done on a tight breaking-news deadline. I have no patience for slow-loading search results.

• Duck Duck Go has a fantastic privacy policy. This is a big deal. It’s one of the main reasons its creator encourages people to use it.

  Google, the current king of search engines, gets a lot of press for its data aggregation used to tailor ads to users’ interests. Essentially they are scanning your searches, Gmail, online documents, etc. to find out information about you and sell you things. Even though they claim some data is anonymized, such as data gathered through its Chrome browser, we know that when such data is collected at all it is rarely truly anonymous.

  Duck Duck Go’s privacy policy states it doesn’t collect any information about you and doesn’t use cookies unless you save certain settings about the site, like fonts. And you can use URL parameters for those instead.

  Duck Duck Go does log search queries and apparently there’s no time limit at the moment, but the site’s creator says he doesn’t know where they are coming from and they are not linked together by an ID number.

  Compare that to Google, which not only logs your IP address and other information about your computer but also uses cookies with unique numbers for each computer and ties your web search history to your Google account.

  Additionally, some services, such as Google Toolbar’s PageRank, send all the URLs you visit to Google. Google’s search suggestions (on Google.com and in the Chrome browser) send your searches as you type them to Google even before you hit Search.

  HTTP referrers — If I search for “how to bury a body” on Google, and click a result to YouTube (for example), the HTTP referrer tells YouTube that I got to it by searching for that term, along with my IP address. Duck Duck Go (and, to be fair, Google’s new SSL search) hides the referrer.

• In Duck Duck Go, the entire search result is clickable, not just the title. This is a small user-interface detail that makes a big difference.
• Duck Duck Go auto-loads the next page of results. For Google you need a Firefox add-on to do this.
• The Zero Click info boxes are very handy (click here and check out the box of information at the top). Sometimes you just need a little bit of info, a general question answered. These are great for that.
• Customizing fonts is cool. Unnecessary for me personally, because the defaults are just fine (and I like them being bigger than you usually see, by the way), but it’s nice to have the option.
• The favicons next to search results are a nice touch, but sometimes they look messed up (example). That could be the site’s fault, though, for all I know.
• Bang search keywords are interesting. I haven’t used them much but I see potential there.
• Disambiguation pages, a la Wikipedia, are very nice.
• Duck Duck Go supports calculations, phone numbers, ZIP codes, ISBNs, etc.

Cons

• I still have to trust someone. Weinberg says he doesn’t record my personal information, but how do I know for sure? He could be lying. I doubt it, of course, because he seems to have a solid online reputation. If he were an unknown faceless developer I would not trust the site.
• I’m not really sure what the “Information” sites encompass. If this was a little more clear, that would be helpful.
• Duck Duck Go requires Javascript. Google doesn’t. This probably doesn’t affect many people, but some people on strict corporate networks, or fans of NoScript, might have problems.
• There’s no News section like Google’s. Not a big deal, but when I search for something on Google, I often flip over to the news section for more results.

Questions:

• When it comes right down to it, are the search results as good as Google’s? They seem to be so far. But it’s possible I’ll hit a snag where Duck Duck Go can’t find something for me and Google will. It’s hard to say. I still find myself using Google out of habit when I’m moving quickly, but maybe over time that habit will change.

• What’s the business model? There are no sidebar ads, even when I search for something that screams “sell to me”. Weinberg says he’s not making money on DDG right now but might put a few ads up. A previous business of his sold for $10 million, so maybe he is not concerned about making money yet.

  You may say this is none of my business, but it’s relevant because many times when money enters the picture, privacy goes out the window. See also: Facebook. Weinberg will need to have the integrity to resist this.

Summary

Duck Duck Go is fast, it appears to have an ironclad privacy policy and it delivers fantastic search results. I’m going to keep using it and may post a follow-up in a few months once I’ve had enough experience to give a more well-informed opinion.

More info

Weinberg, under the username “yegg,” answers questions on Reddit.

It’s a small piece of Javascript you can embed on your Web site, and it appears to have two main benefits: you can track what text people are copying off your site, and when people paste that text elsewhere, it adds a “read more” note with the URL of the article the text came from.

The benefits of this to Webmasters are obvious:

• you can see what text people most often copy from your site, and
• you can use it to track plagiarists.

When people see your “read more” URL upon pasting the copied text, they are reminded that it might be worthwhile to attribute the information to your site.

One Web site using Tynt Tracer is the Daily Mail. Try it out. Go to a Mail article, copy some text (more than just a few words) and paste it into a text editor. You should see the “read more” note.

Should Web sites using Tynt Tracer inform the reader that they are using it?

Many sites (including this one) use some form of analytics software to monitor what people are reading. Webmasters use this data for a variety of reasons: determining which articles are the most popular, selling advertisements or simply satisfying curiosity.

Most do not overtly tell readers they are using such software. I believe most people these days assume their Web-surfing habits are being watched to some degree.

Tynt Tracer is really just another kind of analytics software, so what’s the hang-up?

When I request and view a Web page, my computer is contacting another person’s computer to retrieve that information, so it seems logical that the activity could be tracked.

But (and perhaps this is naive or old-fashioned) when I copy and paste text on my computer, it feels like an activity confined to my computer. I have already retrieved the Web page from the site’s server; what I do with it after that is nobody’s business.

And to what extent are a single person’s copying and pasting habits monitored by Tynt Tracer?

An easy solution for tech-savvy people is to install NoScript for Firefox and block it because Tynt Tracer is Javascript-based.

But many Web surfers doesn’t know about NoScript, so asking them to install a piece of software to protect their privacy (to whatever extent it may be infringed upon by Tynt Tracer) is a bit much.

I don’t have a solid opinion on this yet. I see what I perceive to be the advantages and disadvantages. But I would love to hear your thoughts. Please leave a comment below!

Further reading:

• Tynt Tracer
• Monitor Content Copying On Your Site Using Tynt Tracer – Makeuseof
• Is Tynt Tracer a Boon for Bloggers? – InternetNews.com