So beyond ranking your site well, what makes this engine cool? :)
SearchMe has a pretty cool interface that rotates snapshots of webpages on a carousel format. The search results are pretty accurate too. Searchme allows users to share searchme's snapshot of your web page and share the snapshot with other people.
However the snapshot is not necessarily the best page of your site. Their snapshots appear to be ancient. So you could lose traffic if it's showing an outdated or inappropriate page.
I just think the interface is pretty cool and I'm ranking well ;-)
Not much surfer visits from them at all at this point in time though.
Here's something that's cool. I noticed they had an outdated snapshot of one of my sites that has been redesigned awhile ago. So I emailed them and received a response the next day. Checked their engine tonight and it was updated.
How's that for cool?
The results aren't perfect, but browsing the snapshots is a useful way to explore the web or find what you want. This is one of the coolest things to happen in search technology since Google.
Room for improvement? Oh yeah. A non-intrusive way to report a crappy result would be useful on several levels. When a bad page carousels in front of a user, a little hot-or-not popup/button/choice available from the page would be cool. You can vote out a page without leaving the SERPs and without disturbing the browsing experience.
The user experience is important. Providing good results and an easy way to participate in creating that positive user experience is good. At the very least you don't show that site for that user. Going macro you can use that data to fine tune the algo, make the user complicit in the results and thus feed the sense of community and sense of ownership that comes with being a part of a community.
It's a very pretty interface, and I like the way it ranks some of my sites too. ;)
On very competitive searches, the algo is obviously not weighting the importance of links as Google does. But it seems to like pages with lots of identical anchor text and very heavy onpage optimizing... the kind that used to work in Yahoo and will now get you in trouble on Google.
It's fascinating to be able to browse through all of those different page designs quickly and see them at a decent size... and this may turn out to be the most useful thing about the engine.
Stacks - As implemented in their "Stacks" feature, this browse feature may become very hot... as it turns into a personal or social bookmarking tool that's explored visually. It allows you to save and share customized stacks of pages (or even videos) you select, to arrange them as you choose, and to email them, post to a blog or to Facebook, etc.
To get to "Stacks," click on Tools and you'll find it. They've got a video on the Stacks page that explains the highlights of the feature in about 3 minutes.
I can imagine this going viral fairly quickly.
PS about Stacks...
I think they could improve the usefulness of Stacks immensely by adding titles and descriptions to each page as it comes up... particularly important in stacks that are primarily text oriented and you can't quite read the content on the page.
They could improve utility even more by allowing user comments to be pasted in. This would allow Stacks to be used as a collaboration tool. Editable comments would make it even more useful.
Giving away my content without the ads that pay for it.
Very nice - good UI and results are not bad at all.
Got potential that one...
Discovered this a few weeks back. Had a positive impression. However I agree that they will get into a lot of trouble if their previews show too much readable information. Copyright issues, etc.
But bravo to them. Refreshing to use after that awful Cull thing we were subject to last month.
Hmm. While watching the "How to" Stacks video, look at the URL of that demo.
That's the sort of URL rewriting that Google was talking about just last week as being unnecessary.
There's redundant "folders" in the URL like /#/ and /1/ and they embedded the session ID in it. The folders contain parameter names including the & and the = signs, so these bits are also redundant: &pi= and &stack= and &ci= and &session= etc.
There's no excuse for that not be any more complicated that this:
and the two parameters still exposed, probably don't need to be done that way anyway, leaving just:
Thanks for the comments about searchme.
We have an algo for determining when to reimage a page but we have about 60 billion pages in our link index with about 2 billion active so we don't always get to your page as frequently as we need to, but we're working on it (1,600 imaging servers now, will add more later this year). One way to make sure we image a page in your site more frequently is to put it in a stack, we image stack pages every few hours.
As for the query rewriting, you're probably right but the strange looking folder names are in there to enable our flash interface to do "back" and "foward", you don't really need them to reference a search or stack, [searchme.com?q=apple...] or [searchme.com...] work fine.
Great ideas about ranking pages and comments on stacks, OK if we implement them?
randy adams, ceo, searchme.com
The one thing I do want to see, along with the title of the page I am seeing in the search results, is the URL of that page.
Noted on URLs for stacks, and comments about rewrites.
If you allow comments on stacks, don't allow the posting of URLs in comments (or add the nofollow attribute to them all) otherwise you'll just get spammed out with link drops.
>>This is one of the coolest things to happen in search technology since Google.
I agree and was ready to snicker at the results but not bad at all. I'm into birds, did my usual bird searches, saw the usual sites there, but actually visited sites that I've passed over because the snippet didn't interest me, but being able to basically see the entire page with legible bird thumbnails & photos made the sale.
Great for quick-shopping competitors too.
RandyAdams, how long has SearchMe been online? If this received the fanfare of the Cuil launch, it would rocket.
SearchMe.com doesn't seem to like me. I visited (or tried to) with a Mac using Safari, and got no further than a blank page showing 'Flash Player Installation' as its Title. Worked OK with Firefox on the same machine. Don't have problems with other Flash sites...
Is it me?
This is way better than Cuil.
What UA does the bot identify as?
I don't see it in the logs, and I don't see those sites in the results.
>>I can imagine this going viral fairly quickly.
Adding some 'webmaster tools' might push that along. ;)
>>I think they could improve the usefulness of Stacks immensely by adding titles and descriptions to each page as it comes up... particularly important in stacks that are primarily text oriented and you can't quite read the content on the page...
I kind of like the rollover snippet at the bottom, but I was thinking how I could plug it into a site. A stack sitemap (automated) would be cool but would need titles/desc to work in most cases.
Skipfactor: When we first launched in April 2008, our relevance wasn't much better that Cuil so we didn't want to get a lot of attention. Since then, we were lucky to have hired Yahoo's top relevance guy and he has improved our relevance by 28% in a couple of months. By our internal DCG5 relevance calculations we are ranked somewhere around MSN but we have some work underway which should get us pretty much on par with Google and Yahoo in a few weeks.
Being able to edit the titles and snippets for pages in a stack is a really great idea, we should implement it asap. In the mean time you can create stacks with your own titles and descriptions by using the "add URL" function under the stacks menu. It's clunky but it works for now.
Our index bot identifies itself as "Charlotte" but our imaging bot (yes, we have 2 bots - bad I know - we have a project underway to integrate them) is built on Mozilla so it will appear as a hit from Mozilla running on Linux.
As for the Safari problem, it appears like you don't have flash 9 loaded in Safari - try loading the latest version of flash in Safari from [adobe.com...]
I was an early adopter of SearchMe, very cool but the problems started to show up quickly.
They index tons of spam sites and the copyright implications are kind of frightening.
Since the searchers can see and read the page they're looking for and it completely cuts the webmaster out of the picture which is a huge problem in terms of ad revenues lost.
Last but not least, I tried SearchMe on my older travel laptop and it's a memory hog. If you don't have the latest machine with a big fat memory it chokes pretty fast so it's not going to work well for people lagging in their computer upgrades.
indrediBILL: Spam comment doesn't add up because we actually have to have more stringent filtering than G or Y because our main index is smaller (2B) - we probably err on the side of too much filtering.
As for the other issues, we have never had a site owner complain bout us listing them, but if we did, we would certainly remove them from our index if they asked us to.
Not sure how old your machine was (what was the make/model, how much memory, etc), but the footprint is pretty small - runs fine native on the iPhone, other mobile platforms, etc.
As for the other issues, we have never had a site owner complain bout us listing them, but if we did, we would certainly remove them from our index if they asked us to.
How many site owners know that there is something to complain about? I only found out about SearchMe from this thread.
I suspect that as it gets better known the complaints will start to come in. Reproducing whole pages in a readable form is going beyond "fair use" when it comes to copyright.
I'm curious how you manage to have complete ( text and all images ) pages from one of my image based sites ..inspite of the fact that the image folders are denied ( all other search engines have always left them alone ) and hotlinking is disabled ..scraping ( ignoring robots .txt ) and or manual copying would seem to be able to generate this kind of "screenshot" overlayed with full page text in a flash player wrapper ..Like piatkow says ..this isnt fair use at all ..my copyright has been royally infringed ..
|Spam comment doesn't add up |
I've seen some real junk sites before that republish the same content over and over and over under different domains and they all show up in the same results. It's really easy to spot in a visual search engine because they used the same page template for every domain name.
|we have never had a site owner complain bout us listing them |
That's because you're not a significant source of traffic yet and you honor robot.txt so some of us already have your site blocked which is why you haven't heard anything ;)
Let's imagine for a moment that fate deals you an incredible hand and suddenly you have 50% of Google's search traffic tomorrow. Now imagine that all those webmasters living off advertising revenue see 50% of their revenue vaporize over night because people can read the content on SearchMe instead of having to click through to their site.
Now imagine your only offer is to remove them from your index, which is no visibility in your site whatsoever which doesn't fix the problem, it only changes the problem as visitors still aren't coming to their site.
The issue is the screen shot itself, they shouldn't be so big, or sharp enough to be completely legible, that people can use your site as a replacement of the original website.
If webmasters start pulling out of your index to avoid that issue, then you're less relevant as a resource and people will look elsewhere.
|Not sure how old your machine was |
I was running Firefox on a old Thinkpad which probably has 256MB of ram. After I had trouble the first time I was watching with the Windows Task Manager running and the more pages SearchMe loaded it just bloated the FF footprint until the hard drive started thrashing. Just tried it again and it bloated up to about 150MB footprint and climbing when it started thrashing and literally brought the machine to a crawl.
Doesn't happen on the newer machines, well the bloat does, but the machines have enough memory to handle it.
Besides, it's a Flash thing, not sure there's much you can do about it.
|I'm curious how you manage to have complete ( text and all images ) pages from one of my image based sites ..inspite of the fact that the image folders are denied |
That's because they're part search engine and part screen shot.
Robots.txt applies to search engines but doesn't apply to browsers which has put them in this murky fuzzy gray area where I would agree with you that the screen shot tool should also honor robots.txt since the purpose of the screen shot in SearchMe is similar to Google's CACHE pages which do honor the NOARCHIVE directive.
Somehow I managed to learn about Searchme since it began and I fall in love with it. I mean, the interface.
But, now that we have the CEO among us, I'd like to ask him:
Why your robot doesn't accept cookies?
In my stats, tag based, which are not based on IP, every page load of your robot looks like a new visitor, which is really not cool at all.
Do you have full screen shots in your gallery?
I've just tried searchme for looking celebrity pictures and never in the websites that appeared on the index the images were that clear that stopped me to go to the sources. Or may be you just have a really large screen, mine is 19" and I really had to go to the original source to see the pictures clearly.
Was a "rhetorical" question Bill ;-) I wanted to see how Randy would explain it ..
but yes you are right on the button ..screen shot should honour robots.txt and no archive ..what I really don't like is that this isn't "cache" ( which I dont like either ) this is "serp" ..and a hover on the image gets the searcher all my text ..so why would they visit me ..they have the thumbnails 4 per page about 100x120 each plus each page has a bigger shot of the thumbnails ..on searchme it's nearly a quarter of the actual size on the site ..
this disregard for my copyright makes G look positively angelic ..and no Randy ..I'm not content with an after the act "we can remove you from our index" .."opt out" ..you have to ask first !..your model is "scraper" ..just like the ones out of the east
|Reproducing whole pages in a readable form... |
|The issue is the screen shot itself, they shouldn't be so big, or sharp enough to be completely legible, that people can use your site as a replacement of the original website. |
Are you sure? I did a search of my sites and they're only showing a snapshot of the first few paragraphs of content. I did a search for NYTimes and could only see a partial snapshot of their page. It looks like they're using the first 600 pixels of a page.
It's clear that searchme cannot serve as a surrogate for your site. I would have been the first to complain if that were the case. My first reaction to searchme was a knee-jerk and I wrote about copyright infringement issues. But then I thought to look closer and realized those fears are unfounded, it was just a knee jerk reaction. Bill, the issue is not the screenshot because it definitely cannot be used as a replacement for surfing a site.
If someone wants to actually read your site, they have to visit it. No hit on your AdSense ads, affiliate links, etc.
That searchme allows people to share the stacks with friends is a bonus for webmasters because friend referrals are a great validation for a site.
[edited by: martinibuster at 6:26 pm (utc) on Sep. 28, 2008]
MB ..my images are showing up at 20% of actual size ..the pages are only of 4 thumbs each on the left side plus one large image on the right side ..they have only one paragraph of text ..and the hover on searchme shows all the text ..they are showing the larger image at around 200 pix high ..no need to visit the site ..and they damn well scraped them straight into serp ..and an offer to take them out after the fact isn't going to cut it ..they are out of line ..
They're "image" search is totally broken btw :(
I've just tried the famous female celebrity with a lot of children and it returned me another unknown woman instead.
|Was a "rhetorical" question Bill ;-) I wanted to see how Randy would explain it .. |
If it was rhetorical then why would Randy explain it? LOL
|I did a search of my sites and they're only showing a snapshot of the first few paragraphs of content. |
My point is those paragraphs are enough to either provide the viewer with enough of the content they want or indicate your site wasn't what they wanted and in either case you lost the visitor, lost the page impressions (useful for a media kit) and your AdSense and affiliate ads never had a chance unless the visitor decided to click through to your site to read the rest of the text.
| This 102 message thread spans 4 pages: 102 (  2 3 4 ) > > |