Posted by: Paraic | July 29, 2008

Is that Cuil or what?

There’s already been a lot of coverage of the new search engine Cuil, especially given the backgrounds of its management team.  I wasn’t planning originally to post on it since I didn’t think I’d have anything to add to what has already been written, but from a combination of surprise (not in a good way) and frustration, here are some observations:

The first major claim of differentiation for Cuil is the 120 Billion page index they’re using.  The claim of course is that with all these pages indexed you’re going to be able to find more/better results, especially for less frequent query terms.  But my first impression is that all those extra pages they’ve indexed are not really adding to the value of the results – if anything, after a certain point it seems to be that it’s just noise/junk that’s being added in to the index.  Think about how you get all these extra pages into a search index.  These aren’t the easy ‘static’ pages you can just crawl for content; everybody has those.  To get to this number you try to generate as many variations of dynamically-generated pages as possible.  If you over-do it though, you end up with junk pages in your index.

Take my sample search for CNGL.  One of the results at Cuil comes from a site that helps you find UK company names:  If you browse their directory, you can find companies starting with ‘C’, then those with ‘CN’, then drill down to ‘CNG’, at which point you’ll see your options are now limited to companies that start with ‘CNGA’, ‘CNGT’, or ‘CNG-(space)’.  The Cuil results list though has a page where they’ve attempted to find companies with ‘CNGL’, generating the page with content that says, ” Sorry, no joy looking for that company, try again though…”  (  A similar example is a set of results from a ‘TypoTrap’ at where Cuil results include, “Did you say… CNBL?” and “Oops… VNGL“.  So in cases like this, all those extra pages in the index are not really helping me – and the CNGL home page was not in the result set at all.

My second observation is on the result presentation.  I’m all for exploring alternatives to the standard ranked list of 10 results with snippets (I’m continuing to use SearchMe and still liking it so far), but I found the Cuil results presentation just really disorienting.  It’s not at all clear what the ordering of results is, or even whether there is any ordering of the results presented.  I know that the assessment of which result suits me best is up to me, but I find now that I’m now “trained” from my use of search that there will be a rank ordering of results and the obvious strategy is to start at the first/best and work my way down the list.  The Cuil presentation just breaks that model and it’s disorienting in a big way.  (On the other hand it’s interesting to discover how deeply that model is ingrained – we really rely on the search engine to guide us in navigating the result set).

It will be interesting to see how Cuil adapts and evolves.  I’ve seen some people saying that they’ve already seen changes/updates in Cuil’s behaviour in terms of search performance after just the first day.  I’m sure it will evolve quickly.  Worth watching.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s


%d bloggers like this: