My December 30-days challenge was:
"Use DuckDuckGo as favourite search engine".
It has been an utter failure.
Let me recap the reasons:
- I really missed the "Google DDG Suggest" feature.
DDG makes a point of not tracking user behaviour, and this probably applies too to non-personal query data. To build an auto-suggest feature like Google does you need at least to store all-users query, so if DDG doesn't keep them, it probably has not way to offer it.
Google Suggest causes a lot of pain (lawsuits, complains, angry Romanians, etc...) but I never realized how dependent on it I was. For example, whenever I have doubts about the spelling of an English word I start typing it in the Google search box; it's not the search feature I need, it just happen to be quicker than consulting a dictionary. DDG doesn't help me there.
Another example is when I misspell a term in the query, before submitting it Google suggests me the correct spelling.
These two examples could be easily achieved without storing user queries, you just need a dictionary-based suggestion.
Google Suggest is what I missed most
- I missed search integration with Image Search. I don't use Google Images that much, but the results from Google Images in the web SERP are really useful; DuckDuckGo should probably consider a partnership with a search image provider (TinEye?)
Google SERP gives a visual hint of what a waterfall is
- My personal mailbox is a GMail account, I started recently exploring and using more seriously Google+, I leverage Google Docs Drive... all these tools are tightly integrated in the Google ecosystem and it's easy find yourself staying within its boundaries.
Now, let's be clear: I like DDG, I really like the way it presents results, its goodies, how discrete it is with advertising (remember how G used to be?).
The fact is nowadays user expectations are high, and to win the number guy you don't just have to win, you need to surclass in order to unlock users from their habits.
All this said, "utter failure" would sound too ungenerous. What made the verdict so bad was not DDG's fault, and the verdict is not about DDG performances - fairly good indeed - It was more about my challenge.
The factors independent by DDG which made my 30-days challenge fail are:
- The DDG for Chrome extension caused instability in Chrome; it took me some time to locate the culprit, but made me somehow start to estrange from DDG (I'm giving it another spin now, at least I'll be able to report an issue to the DDG team if it has not been fixed already)
- I work too much with Search/Google, so I had to add a clause "for non Google-specific searches" (e.g. site:, etc...). The clause seemed to apply to far too many cases
- Whenever I saw a SERP in DDG, I felt compelled to compare it with Google SERP. This is especially true for Italian queries, where I had a little bias (often wrongly put) on its abilities.
Not that I wasn't overall satisfied with its quality, generally speaking it was quite good for the first search results, but Google is too much entangled with my mind as "the reference".
What I dislike most of Google is the abuse of ads on its SERP. Nevertheless, I admit that for the majority of queries I enter - programming related, or general knowledge questions - Google SERP is not cluttered with ads (usually there are not ads at all, while DDG shows always one - non disturbing - at the top).
The response time is usually a little better with Google. Again, DDG does a decent job considering it has far less resources (and traffic) and still depends a lot on third party data providers, but every tenth of second adds up in user experience.
What makes in my opinion DDG a better user experience is the case of commercial related queries, where 90% of Google SERP above the fold is made of ads. Professional life is different, but as a user I don't search them much.
Google's position is that ads are relevant to the search query, and might be even true, but usually I look for general information on a good before looking for a place to buy it.
For now, I will stick my habits to Google Search. I probably am not a normal user, and really believe DuckDuckGo could be a good option in the future; it has much room for improvements and I think it should work hard to emerge and conquer a larger user base.
An initial draft of this article was originally published on my G+ page, and I mentioned it on Twitter to DuckDuckGo.
To my surprise, it didn't go unnoticed, and DuckDuckGo replied on Twitter in a very short time:
Listening to criticism is something that increases the chances to succeed, and definitively impressed me. I'm looking forward to test the new feature!
Update 28 January 2013: DuckDuckGo replied to this article on Reddit, once again demonstrating great attention toward user feedback; they also anticipated something more :-)
And you, dear reader, haven't you tried DuckDuckGo yet?