Cuil - the New Search Engine - Issues
There’s been a lot in the news this week (so far) about a new search engine called Cuil. Apparently it’s a well-funded start-up of educational types and former Google employees who are looking to do better than Google itself. That’s a tough mountain to mount, given that Cuil wants to unseat a company whose name is a verb that describes what Cuil offers: search.
Cuil does not offer the many ancillary services/products provided by Google, such as image searches, local searches, map searches, online word processors, and email accounts. What Cuil focuses on is just search. It claims to do a better job of offering more relevant search results by ranking pages’ content and relevance, rather than inbound links - what Cuil refers to as “superficial popularity metrics”, and which Google uses as at least par of its search algorithm known as PageRank.
According to their home page (as of right now) Cuil also says that they have managed to index 121,617,892,992 web pages, compared to Google’s reported 40 billion.
That’s all well and good, but none of that provides an answer to the most basic question that defines how well a search engine works: does it give me what I am looking for?
To that end, I ran a few searches and got mixed results. Searching for Jonathan Schellack did not return this web site. The search returned 17,986 results, but only showed me a subset of that on three pages. Where are the rest, I wonder? In none of the visible results was www.schellack.net/jonathan.
Searching for yourself on the web may be the first thing we self-absorbed people do, but how about the things I search for on a daily basis (I promise I don’t search for myself that often). So I tried out a search for something I had used Google for several days earlier. I entered “sql cumulative returns” to try and find some examples of SQL statements or scripts that would show me cumulative, aggregate information about some data sets I was using. My first search returned entirely disappointing results:

Thankfully, refreshing the page gave me better success, actually giving me some information on the second try. Unfortunately, though, the search engine only gave me results that used the word “returns” as a verb, and I had been looking for that word used as a noun. Altogether, the results we not as useful as what I had found on Google.
When I attempted to tell Cuil to give me results with the phrase “cumulative results”, I got nothing (literally). The page was empty of results, though it told me there were plenty out there:

Trying to search for a less obscure and nerdy topic gave me even worse results. I’m considering picking up an iPhone this weekend when my AT&T contract comes up for renewal (more on that later). Of course, I’m hesitant to buy any technology product that I’ll be locked into using for the next two years, so I searched for “iPhone 3G issues“. All but one of the results on the first page from Cuil were from the same, spammy-looking web site. None of them were relevant:

Cuil presents search results in a different and potentially very useful format - and I love not having to scroll down to view my results - but the fledgling search engine still has a ways to go before I’ll consider using it every day.
Posted in web | 4 Comments »


