|Massive Cuil Search Engine Launched|
"World's Largest Search Engine" is the claim
|SAN FRANCISCO — In her two years at Google, Anna Patterson helped design and build some of the pillars of the company’s search engine, including its large index of Web pages and some of the formulas it uses for ranking search results. |
Skip to next paragraph
The makers of the Cuil search engine say it should provide better results and show them in a more attractive manner.
Now, along with her husband, Tom Costello, and a few other Google alumni, she is trying to upstage her former employer.
On Monday, their company, Cuil, is unveiling a search engine that they promise will be more comprehensive than Google’s and that they hope will give its users more relevant results.
Quick recap of 10 pages so far:
1. The results suck
2. The domain name sucks
3. The image thing sucks (and is a lawsuit waiting to happen)
For me, I'm struck by the reaction to the layout. Some strong opinions against and yet others in favour. I really think they are on to something. Leaving aside the issue of the quality of the result, the way that users experience this tool is very different to the Google experience. It reminds me of the way men and women go shopping:
Man knows what he wants before entering store; looks through aisles; if not in first aisle, try second; not in second, try third, etc. Keep moving. Do not ask for help or directions. Get product, get through till, get the hell out.
A woman might even stand still in a store. She knows vaguely what she wants, but not exactly. (eg. Something to wear at the weekend, but will it be a dress, a trouser suit, blouse/skirt, jeans/t-shirt?). A woman takes suggestions from the merchandising layout, the mannequin examples, she browses the rails, she might even ask the shop assistant for help. It's still goal-orientated, but not as a man might understand it.
Searching on Google, you try the first phrase. If results don't fit, add another keyword. Still don't fit, add another word or change one. Remember, Google has said that over 50% of all searches are completely unique. One interpretation of this is that users try several increasingly specific phrases before finding the results they want.
Cuil begins with the premise that the first phrase a user enters will not be the one that returns their desired result. The drop-down menu immediately suggests alternative, more refined phrases. The results are not ordered hierachically, they allow the user to browse more freely; the Explore By Category feature offers still further suggestions. Its a winnowing process that is very different to the get-in-get-out Google experience. Looking at Google's 'searches related to ...' feature at the foot of the SERPs, compared to the Cuil experience, it almost seems like an afterthought.
Two other thoughts about the Cuil results layout - it messes up all those SEO reports that require a hirarchical list; and it works well on modern higher resolution screens (more results above the fold).
If only Cuil could get the results right, the difference could well provide a viable alternative.
quiet_man, thank you so much for the thoughtful analysis and the excellent points made.
I believe there's some solid reasoning behind the layout, since searchers do refine their searches down to more specific ones, and if they go to a site on a more general search term, they still have to either use that site's search or drill down in the navigation to the specifics they're looking for. I think the interface (while it does need refining) reflects a grasp of user needs and how people surf - and shop.
I don't think I can possibly express how badly I feel about this launch and the press and public reaction to it, empathizing with how the people must feel who have worked so hard on it. I realize people are being honest and candid, but emotionally it all just has to feel so hurtful to the people who have worked so hard on this project.
I can just imagine how it must feel in the hearts of the people involved, but there's some solid talent there and I sure hope they can debug their creation and make it functionally feasible.
|there's some solid talent there |
That "solid talent" was not bright enough to spot the most basic problems, all mentioned after a quick glance at the service. So I guess their talent is not as solid as their job histories may suggest. I mean, haven't they heard of "focus group testing"? Haven't they understood the concept of slowly upscaling a system by using closed alpha/beta testing groups? And how comes that that image thing is not fixed yet?
Cuil still looks more like a research project to me instead of a commercial service (even a buggy one). It could also be just a hoax, designed to put Google's position into concrete.
|I don't think I can possibly express how badly I feel about this launch and the press and public reaction to it, empathizing with how the people must feel who have worked so hard on it. |
Marcia, I am not an unsympathetic person but I must disagree with you on this. They have evidently not worked hard enough. Their search engine is a joke on many fronts. Surely they had enough intellect in there to test it before going public?
This must be one of the worst launches ever, worse even than Vista. Cuil is the Ford Edzell of search so any sympathy is perhaps a wee bit misplaced.
Has anyone from Cuil joined this forum yet to answer these criticisms or are they all in hiding?
Conspiracy theories aside :) - I agree with Marcia and Zett; the 'founders' have had pretty strong reviews from their old pals at Google, and from outside; there's no doubt they are highly able folk.
On the other hand, what we've seen is nothing for them to be proud of.
When the dust settles, we'll almost certainly find that the launch exaggerations and half-truths were the result of giving too much rope to the marketing department, a common error these days, and the launch before sufficient testing was almost certainly down to shareholder pressure - a common occurrence down the ages.
I find it incomprehensible that anyone with a major web project - but with such powerful competition - has not learned the secrets of viral marketing and intelligent understatement.
Any company competing with Google who even mentions Google in the product's first year - let alone claims to be bigger, or better in any way whatsoever - needs to be 1000% sure that they can *prove it*, without one single flaw, from Day 1.
Anything less isn't just suicidal, it's plain old fashioned stupid.
And their only hope of a second chance now, is to say We Got It Wrong.
[edited by: Quadrille at 11:46 am (utc) on July 30, 2008]
Excellent summary quiet_man.
It needs time and a few iterations but personally I like the combination of search engine and directory, which is how I see it.
Marcia: " I realize people are being honest and candid, but emotionally it all just has to feel so hurtful to the people who have worked so hard on this project."
Peer review is always scary, and then you always get bashed as well with almost every project, and this goes for almost everyone. The Q&A session after the presentation at a conference for example is way scarier and more difficult than the presentation itself. We're the users, giving honest opinions, without trying to finger point at individuals in the team. As scientists they shouldn't take it personally, they're technical issues...things go wrong sometimes.
For a giggle, I've had amongst others said to me "Your work is boring", "I can't understand a word you're saying", "why didn't you test all these methods?" (unknown to me), "why doesn't it work properly?", "You have a lot more work to do", etc...but i have had a lot of good comments too :)
Bottom line, I learnt a lot from that criticism, although it was horrid to recieve in a crowded lecture theatre. Humbling.
User review is really important, and I'm sure they're getting loads from it. It's great that people are talking about it technically, although from a business perspective I'm not so sure it's good for them.
[edited by: Misscj at 11:56 am (utc) on July 30, 2008]
|pageoneresults wrote: |
The name Twiceler became a household name in the Bot Blocking Communities. And now "I" know why. It was nothing but a scraper.
Yep, been blocked via .htaccess files for a looooooong time. They need to invest some time, money, research in a crawler/bot before even thinking of becoming a SE. Pretty sad really...
I just discovered something interesting:
I tried a unique/distinctive 5 character string (coined term) that is the domain name of an obscure domain registration site, and on thousands of domains that have been parked by this service.
The result for this search is an entirely blank page, with this text in the upper right corner: "2,568 results for -----" The rest of the page is entirely blank -- they don't display any of those thousands of results, including -----.com.
In contrast, Google displays "Results 1 - 10 of about 2,540 for -----. (0.21 seconds)at the top of the page, and then provides numerous pages of ordinary looking search results.
The -----.com site itself is midway down the first page; most of the rest of the results seem to be spam of one sort or another -- mostly junk generated by scrapers that have auto-incorporated this distinctive string into pseudo sentences, etc.
|The result for this search is an entirely blank page, with this text in the upper right corner: "2,568 results for -----" The rest of the page is entirely blank -- they don't display any of those thousands of results, including -----.com. |
I had the same thing only number of supposedly existing results were measured in millions. This is either all servers not working correctly or (more likely in my view) they have a very, ummm, creative way of counting "indexed pages" and consequently "results".
|I don't think Cuil is as bad as some make out, but they've clearly shot themselves squarely in the foot with the worst error-checking I've ever seen on a newly-launched site. |
That's good that there are those who have a balanced perspective on this. Do some more searches then come back and give us some more feedback. The 100+ searches I've performed are more than adequate to provide me with an overall picture of Cuil.
|Overall, the results are not that bad though. Not good enough, but not terrible. |
No, not terrible and surely not good. Yahoo! and Live return better.
|Cuil may well have a bigger index, for what it's worth. |
RA, how would Cuil have a bigger index? I mean, Google has been indexing the web since when? Google have over one million servers to crunch data. What does Cuil have? Two VMs on a shared IP at GoDaddy. < Just kidding but it seems that way.
No, I don't even think their index is bigger than Gigablast which has been indexing now for more than a few years.
|The list of results is a rather prehistoric approach to displaying results and something else is needed as technology in SE's starts using more and more A.I and so forth, meaning that ranking as we know it can't really work that way anymore as they become more intuitive, asking the user for feedback and making those results more flexible, giving more space for accurate results to suface for you. |
If anyone is going to test a new layout for SERPs, it would most likely be Google. They've set the standard for "regular search". Trying to bring a "new" layout to the public that is totally counter-intuitive to what we've used for years is a rather lofty goal.
|I mean if cuil worked better than Google, I would bet that the layout wouldn't matter so much anymore. |
If. Ya, if Cuil worked better. Its unfortunate because their less than satisfactory search results are trumping any other positive thing they have going for them.
|We know from our research that people can make better and quicker decisions about relevance and quality when they can see an image from the website. We do our best to take images from Web pages that accurately reflect the content of the website. Many websites are full of images, so we use advanced algorithms to determine the best image to show the user. |
The absolute worse failure I've seen to date. When I perform a search for a well known name, I'm expecting to find pictures of that well known name somewhere in those first results. I'm not expecting to see a babe in a bikini on the beach oiling herself up. No, that doesn't work.
|Apparently they have been accidentally inserting adult images into search results |
Accidentally? Nah, that's the algo at work.
|1. The results suck |
2. The domain name sucks
3. The image thing sucks (and is a lawsuit waiting to happen)
Good 3 point summary!
|I don't think I can possibly express how badly I feel about this launch and the press and public reaction to it, empathizing with how the people must feel who have worked so hard on it. |
At this point, I no longer feel anything for them other than disgust. Every single statement I've read since launch has not been in their favor. Ya, I'm surrounded by SEO types but that should be the real test. If anyone knows search, we do. Screw them, they deserve everything they have coming at them right now.
|I realize people are being honest and candid, but emotionally it all just has to feel so hurtful to the people who have worked so hard on this project. |
They'll get over it once they move on to their next job. Look at it this way, your Cuil experience is something you can put on your resume moving forward. Oh wait, you may want to leave that time frame blank and just say you were taking time off due to personal reasons.
|I can just imagine how it must feel in the hearts of the people involved, but there's some solid talent there and I sure hope they can debug their creation and make it functionally feasible. |
Must not be too solid if they left Google. And, Google let them go. Makes you wonder doesn't it?
|It could also be just a hoax, designed to put Google's position into concrete. |
Ha! Now there's another angle.
|Their search engine is a joke on many fronts. Surely they had enough intellect in there to test it before going public? |
And this seems to be the common sentiment amongst many in our industry.
|This must be one of the worst launches ever, worse even than Vista. Cuil is the Ford Edzell of search so any sympathy is perhaps a wee bit misplaced. |
My sympathy goes out to the top three right now.
|Has anyone from Cuil joined this forum yet to answer these criticisms or are they all in hiding? |
They would be best served to remain in hiding at this point. No matter what they say, or do, it is too late I think.
|On the other hand, what we've seen is nothing for them to be proud of. |
No it is not. And, that reflects on their abilities.
|Anything less isn't just suicidal, it's plain old fashioned stupid. |
But wait, you just said above that they had talent?
|And their only hope of a second chance now, is to say We Got It Wrong. |
Nope! That chance was gone a few days after launch. Now they are just wallowing in it.
|Yep, been blocked via .htaccess files for a looooooong time. They need to invest some time, money, research in a crawler/bot before even thinking of becoming a SE. Pretty sad really. |
The writing was on the wall when their bot was out traversing the web and not being nice. Those who delve in that area quickly caught on and blocked the pesky bugger. Those who don't, now have their stuff indexed by Cuil and what a mess that is.
Okay, how many more replies before Cuil becomes yesterday's news?
|how many more replies before Cuil becomes yesterday's news? |
I would like to explore the angle on the WHY.
WHY did they ignore most of the basic rules when launching a product (let alone a product that was spun as "Google competitor")?
I have seen very few answers to that question so far.
1) Google. Conspiracy theories on Google backing this failure to avoid future PR blitzes by other startups that might really turn out to be a competitor.
2) Financial power. After having burned money for about two years, investors forced them to launch, no matter what. If VCs did this, they deserve that their investment vaporized on Monday with the "launch". (And no, I do not have sympathy with the VCs, because they should have/could have detected the failure long before.)
3) Lucky losers. They are just a group of not-so-bright folks who were lucky. Lucky in the good old days, lucky in somehow making it to the Plex, lucky in being able to beef up their CVs with their time at Google. This made a good impression in their CVs. Probably Google also added a bit of financial independence to their lives. And this was extremely helpful convincing VCs to shell out $33M. Money that -given their limited abilities to launch commercial products- was not invested wisely, but that's the way VCs are.
4) Hubris. They honestly believed that they could not be wounded. After all, they WERE coming from Google. They HAD the experience from other projects before. They HAD some ideas to revolutionize web search. So who was going to stop them? And they behaved like Google always did (mostly ignoring copyrights, putting out alpha/beta products to the open market, keeping the whole operations in the dark). Then they suddenly realized that Google can get away with this, but they -as a startup- can't. The press and industry specialists can be ruthless, especially if they have been teased by a "Google Killer" story that they re-published unchecked.
5) Sales pitch. During their development they realized that they could not compete with Google. But some parts of their business were still valuable (the crawler, maybe, or the design team) and the "launch" would attract enough interest for a cheap, quick acquisition that at least recovers the VC investment.
Given the massive failure across the board, I would really explore the WHY. It's interesting, because we can certainly learn from this failure.
If they are as bright as they want us to believe they are, they must have known that their service is crap. So why did they launch anyway?
If, on the other hand, they did not realize that their service is crap, then they are by far not as good as their CVs suggests.
|But wait, you just said above that they had talent? |
I think it's quite important to distinguish between the individuals and the company; context is all.
I think they have very bright people; but the whole may be less than the sum of the parts.
It's a well-known feature of organizations! :)
|Okay, how many more replies before Cuil becomes yesterday's news? |
It ain't over until [political incorrect thought self-censored]
It will fade, as they all do. But I suspect they'll be back. Hopefully with an honest relaunch and realistic objectives.
It's a cruel, hard world out there, isn't it? Makes me glad I'm sitting here in my dining room wearing shorts and no shoes this morning, instead of having to go out and face corporate charades and facades and the mean-spirited people of this world.
IMHO, what Quadrille said is spot on (even though he's a white hat [sic]). They're smack in the middle of the public spotlight as a result of some atrocious PR and marketing. Now they need to put some flesh and bones on the technology and the hype, and come out in public and show their faces to the public.
"We goofed and we're working on fixing it" sounds like a good start.
Some of us are on their side and would like to see them succeed with something innovative that works. Someone needs to shout, "C'mon Cuil.com, come on out and talk to us!"
|RA, how would Cuil have a bigger index? |
It's important to make a quality distinction - the one that Google also avoided making too prominent with their announcement about a trillion URLs. All that's needed is a random string generator and a local server to index more than any search engine in existence. I think the 'largest index' claim was foolhardy and disingenuous, but potentially it is factually accurate.
|If anyone is going to test a new layout for SERPs, it would most likely be Google |
One advantage any start-up search engine has is that Google are scared of major change. Quite reasonably so, since users like the Google format, but I see this as a weakness that could potentially be exploited by a search engine with little to lose.
As for the results, I think they're much more interesting for generics than more specific searches. It isn't for me at all, since I search all but exclusively with advanced search operators, but I think they're mostly results than a typical end user could live with. Relevancy is an ongoing process, and I wouldn't expect anyone to get it right for every search from day one.
I confess I'm surprised at the marketing behind the launch though, which seems to be an unmitigated disaster. Too much too soon. A 'stealth' launch would probably have attracted a much more favourable response from influential groups. Instead, they've set themselves up for widespread ridicule.
This viral news cycle is due to the high hopes that so many had for this launch -- not as a "Google killer" (what a dumb phrase!), but rather, as a strong competitor.
Whatever the reason their top people left Google, the fact is they DO have impressive credentials. But however bright and experienced they may be, they made unbelievably elementary mistakes.
It would have been a lot smarter to under-promise, then over-deliver. They did the opposite, which set them up for the fall.
In the world of carpentry, the rule is "measure twice, cut once". I don't get a sense that the Cuil folks measured at all (how could they possibly miss the wrong images next to almost ALL the results?).
No one wants another major search engine player more than me, and I am willing to cut them some slack in the short term, but I must say, this is a real disappointment. The potential was there, but it was not delivered.
[edited by: Reno at 3:07 pm (utc) on July 30, 2008]
I think I know why Cuil was launched. Patents. If there's one thing Ms. Patterson knows it's getting patents granted for search engine algorithms. Maybe they just want to patent a pile of techno-speak before Google gets the chance to. The patents will be easier to enforce if they are actually used in a product. If the patented algo's actually work and the patent possibly hamstrings future google changes, then they can sell the rights to Google.
They'll probably make a lot more money selling their newly patented ideas back to Google than they would as Google employees.
Zamboni, you don't have to have a patented method in a commercial or even working system for that matter. It has to be original and you have to prove that basically. So I don't think it's for patents that this was conjured up.
[edited by: Misscj at 2:58 pm (utc) on July 30, 2008]
At this point they should be offering webmasters the chance to present their own "Cuil Favicon" for the listings, through a tag.
<link rel="shortcut icon" href="cuilicon.ico" type="image/x-icon" />
That way people can throw up their logos and whatnot. Then they can use their non-working algorithm for everything else.
It would be a good move on marketing's part anyway to extend an olive branch for the disaster they've already caused. I would jump on that offer immediately just to get other people's images out of my listing.
It's better than a lawsuit anyway, which is what is bound to be on the horizon if not fixed quickly.
|At this point they should be offering webmasters the chance to present their own "Cuil Favicon" for the listings, through a tag. |
I personally think this one will go places if they can only get better results. I'm seeing a lot of duplicates and irrelevant sites. Plus my site gets a different logo slapped up beside it instead of its own.
But the "browsing" format and features of the SERPS could potentially be hugely popular - I know I like it. I suspect this will be a big hit with people who don't know how to narrow their search very well.
However, the usability is very poor for going beyond the first page of results. I didn't even notice that black bar down there until I'd already done several searches; it just doesn't look like part of the viewport.
|"We goofed and we're working on fixing it" sounds like a good start. |
That is absolutely the best advice for these people, they have got to get out in front of this thing right now, and right here wouldn’t be that bad of a place to start; “This is what our vision of this thing is, and here is why it went wrong, but were excited about getting that fixed”.
I do feel a bit for them, had some launches of things go badly myself so we have all been there.
I want to be encouraging and not negative but some of the mistakes they have made are just mind boggling; overloaded platform, black interface, name that’s confusing, and they must have a stack of letters from lawyers already on this image problem. I have seen major companies with logo’s swapped.
With all of that said, I still believe these guys have something very significant, and the train wreck looks a lot worse than it really is. Trouble is people will give you some time, but in these days not a whole heck of a lot before the only place they can find you is here; [archive.org...] the bone yard.
--Two other thoughts about the Cuil results layout ... and it works well on modern higher resolution screens (more results above the fold).--
But that's also a problem, a lot of people now use search on devices which don't have high resolution screens, such as mini-laptops or smartphones. In a few years time it may be that the majority of search is conducted on portable devices of one sort or another.
At least the old-fashioned list approach is flexible enough to work on absolutely any size or shape of screen.
--If only Cuil could get the results right--
...and the images. The images do not seem to fit the links at all, they're frequently from a rival site, or sometimes not even to do with the topic at all.
What do people think of the explore by category box that you get on some searches? I found it vaguely intriguing to start with, then the penny dropped that they were simply taken straight from Wikipedia categories and there seemed to be little connecting them.
It seems to me that the Cuil algo is capable of looking through hundreds of things within a given category and finding a handful of the least relevant ones to your search.
|Some of us are on their side and would like to see them succeed with something innovative that works. |
I 100% agree. Some real competition for Google would be a great thing for most of us. But they have blown their chance big time.
Yes, they've blown the chance with inadequate preparation and testing and really BAD marketing. But now they need to ride the wave of the massive exposure they've gotten by making a public appearance, starting to actually communicate, and learning how to "work the crowd."
It irks me no end, but I'm afraid I see a lot of similarities between the Hollywood/Entertainment/Recording Industry's public BS and buzz and the search industry's BS and buzz. It always irked me with that, and now it irks me with this industry.
There has to be some room left for openly and honestly communicating at the public level, and IMHO that's exactly what these people need to do right now.
|But they have blown their chance big time. |
No they haven't.
It doesn't matter two hoots what webmasters think. If they start getting some traction and traffic we'll all be bending over.
|It doesn't matter two hoots what webmasters think. |
Heh, that's what AltaVista thought. ;)
>>But they have blown their chance big time.
>>No they haven't.
So is it possible that maybe they haven't? When you think about it, mediocrity would have forever doomed them to oblivion as just another also_ran. Ho-hum. But this fiasco has elevated them to the publicity and exposure status of "Brittney goes back into rehab!"
It's a sad fact that in show biz, politics and search, it's just about all come down to the point of knowing how to "work the crowd." And as a sidenote, we all know who does that sublimely well for Google, don't we?
|If they start getting some traction and traffic we'll all be bending over. |
Do you think they will then?
|Do you think they will then? |
My crystal ball is no better than yours.
But if they do, I will allow them to crawl and send me traffic. The real acid test is not us, but users. Who cares what we think. Certainly not an SE. Cuil are currently laughing at our comments. They believe they have what it takes and I wish them the best of luck.