Tailrank founder Kevin Burton notified us that version 2.5 of his news aggregation site has launched, as well as a new version of the engine behind it called Spinn3r. We’ve taken a look at the new site, and in our opinion it still falls short of being a useful application.
We’ve been a bit harsh on Tailrank over the last few months, even suggesting that it may be time to deadpool it. But the site was without any content at all for a few weeks, and when Burton said it was fixed the site was filled with spam (Burton writes about the spam attack here).
So back to the new version…the spam is gone, but the stories are all at least a day old. Burton originally promised this release in early July. It came three months later, which is not unexpected when software is involved. But he knew that we’d be taking a critical look at the site. If his indexing engine can’t keep up with the news, how can he expect people to spend time visiting the site? We just criticized competitor shoutingmat.ch yesterday for the same problem. This is a competitive space (Techmeme is the clear leader, and there are lots of others), and anything short of perfect won’t stand a chance.
We’ll keep giving Tailrank the benefit of the doubt and hope to see it improve soon. But I’m not sure anyone else out there will do the same.
Mike,
It’s not old news… it’s stable news. People are still talking about these stories.
For example, the top story:
“Bush vetoes expansion of kids’ health insurance program”
has nearly 900 posts. It had about 600 this morning.
If you want the latest and greatest you can click on the Tailrank river view.
http://tailrank.com/river
The stories here are less than an hour old.
I think some of your criticism is somewhat valid though. We’ve heard from a number of people that they want more ‘velocity’ in the result set.
This is actually a pretty trivial dial for us so we might be changing it in the next release to see what our user base thinks.
There are some more UI features we’re thinking about though that will make 3.0 work better with content that changes with high velocity.
Onward!
Kevin I don’t think you should display so many posts for each story. For example, the Bush story you mention… There’s almost 20 related stories for that shown by default, which is simply too much. I have to scroll down like 4 pages just to get to the next story. I don’t like TechMeme personally but I think they have it right where they only show one or two related stories, which is enough.
Well, best of luck with your site.
Kevin,
I think the win here is to recognize stable news, like you said, but way earlier in the cycle like techmeme does. Thoughts?
i personally like techmeme.
yeah I agree with Andrew techmeme is better.
http://vidsonly.blogspot.com
“has nearly 900 posts. It had about 600 this morning.”
I think thats the problem with tailrank, its doing what its name implies. ie. ranking the tail end bloggers who say look at the news once a day.
The reason techmeme picks stuff up so quickly is because its only tracking a couple of thousand very prolific blogs.
Techmeme is getting suckier, though…. when/if it falls into the cesspool it would be good to have something in the wings to replace it.
threz,
One of the reasons that Tailrank tracks millions of sites is because I believe that this is more democratic than only tracking thousands.
Google news tracks thousands? So does yahoo news.
There are other ways to track the cool stuff other than ditching democracy
Kevin
Mike,
“I think the win here is to recognize stable news, like you said, but way earlier in the cycle like techmeme does. Thoughts?”
Maybe not so much as recognizing stable as recognizing HOT news.
Some of this stuff will make it into 3.0…… there have been features and algorithm tweaks that we’ve been mulling over for a year that should make it there.
Great constructive criticism.
Thanks guys.
I have been to the site and I think the amount of scrolling one has to do to go from story to story means it will be a long time before I go there again. Maybe 3.0 will fix that.
Techmeme is pretty sweet.
Their homepage is so damn long and my firefox freezes to death.
I think, they should be doing some filtering for homepage and display less important or previously popular ones in a different way.
“One of the reasons that Tailrank tracks millions of sites is because I believe that this is more democratic than only tracking thousands.”
With all due respect Kevin, I think aggregators and more suited to meritocracy than democracy. Aggregators owe nothing to sites in their scraping loop, they owe everything to readers who want distilled lists.
Paul,
Trust metrics (like Tailrank) can exhibit both meritocracy *and* democracy. For the record we don’t consider all blogs of equal value.
“Aggregators owe nothing to sites in their scraping loop, they owe everything to readers who want distilled lists.”
I tried registring but got an error message instead:(
Fair play to Kevin Burton for being a proper, never-say-die, passionate entrepreneur. Tailrank might suck a bit at the moment, but get ready for version 10.0 – that’ll be spot on.
Users have different needs. Of course most business-users would require news aggregators that pick only the biggest berries and do it quickly. However, there are situations where you are not just interested in what’s news but also how it’s talked about and how far the message is spreading. This could be the case, for instance, if the news is about your own product, service or campaign.
Also, as a researcher, I like systems that try to be as all-encompassing as possible.
Good luck, Tailrank.
New tailrank looks good, big improvement from last one and definitely a useful site.
Kevin,
i have a suggestion.
Why not just use double column ?
One show the main heading with one or two related news, and the other show all the other twenty links for same news but reported by different site.
Kevin,
I do not think one have to be overly religious in making sure that millions of site is been tracked instead of thousand of site.
What is important to user like me is quality not quantity. I think you start from the wrong perception that user care more about quantity than quality. I say ‘hell no’. Why ? Because more or less, the news and words for similar news are almost the same. The main thing is to convey news to us, that what really matter. We don’t need millions of information to know what happen because we have other news to read on.
The success of Techmeme has been that it is conveying up to date information of a piece of news . Yes, you heard it, just simple a ‘news’ not many similar news’….
Hope Kevin you understand what I trying to tell you… The most important is to help us become knowledgeable and yet prevent information overload to us.
As of current, when I use TailRank, I have feeling that it is giving me more information than I can swallow.
Kevin,
Is there anyway you can tell me what differentiate tailRank from other aggregators ? Hmmm, let me see if I could help to find unique differentiation for TailRank.
When I first saw the site, my first impression was “this is just too much information”. I like the idea of aggregating the hottest news but there must be a way to get that information into a better visual user interface tailored to the average reader.
I believe the key to selling this to the average reader is how fast it delivers the information to the reader visually. I’m not talking the how fast it aggregates news (with its backbone), I’m talking about how fast it catches the attention of the user when you first look at a page. You have to remember that the average reader has a very short attention span. That is why techmeme is great because you can go to the site and glance for the info that interest you quickly. You can’t with Tailrank because you’ll have to sift through 20-30 articles of the same topic (which may bore you) before you get to the next topic. Now if that next topic bores you as well, you’ll have to go through the next 20-30 articles to get to the next topic too. Then I’ll reach the bottom and I see there are 2700+ more articles and it just took me a longer time that I wanted as I had to sift through the articles and topics to see what interest me. I think you get the picture.
That on its own is such a put off and you won’t find many readers coming back just for that fact. It might be good for someone having lots of time to spare but I believe the type of readers using Tailrank (or the type of readers you may want to target) won’t fit into that category. The fast pace and little-time-to-spare techs, reporters, executives that are going to read Tailrank won’t have that luxury.
Personally, I think the first page and inner category pages should only show the title/ topic and how many blog/ site links to that topic. You really don’t need to show the 20-30 of the popular sites/ blogs that write about the topic. If someone is going to be interested in it, they’ll click on the link that would bring them to the summarized page of blog links/ site screenprints. You don’t need to force someone to look at 20-30 posts before seeing the next topic. That way, a reader can glance at the topic and move on if it bores him.
Alternatively, if you really insist on showing the 20-30 blog articles, change the one column layout to a multicolumn one, concise it into a cloud format or something of that sort.
As I said before, the average reader has a short attention span and to scroll down to the bottom and see that there are 2700+ hot articles more is kind of daunting. I am interested to see how many people go past the 2nd or 3rd pages with lists after list of blogs of the same topic.
One other thing you may have to change is how users perceive the site when they first go to it. Michael Arrington and many of us here had the impression that it is an up to date news aggregator like Techmeme. I know understand that it is not, but only after you reading through your comments and the site further. You will lose a lot of readers with those perceptions when they look at a site with a topic that is 1 day old. If you have the ability to aggregate up to date news, you should put it up on the menu or somewhere more prominently. The fact that you had to point someone to http://tailrank.com/river in a comment here on how to find the latest and greatest is not a good way to do it. I think a filter that shows hottest news in the past hour, 1 day, 5 day and 30 days would be a good option too — that way you give the reader more options too.
I see the differentiators, but it is not clearly defined. People are comparing this site to Techmeme and you need to change that.
TailRank shouldn’t complete with Techmeme otherwise TailRank will better off renamed to Tailmeme
Maybe it’s time to drop the whole news thing and turn it into a site where users can post pics of nice behinds (tails) and allow users to rank them.
You guys are spot on in a number of areas:
“What is important to user like me is quality not quantity. I think you start from the wrong perception that user care more about quantity than quality. I say ‘hell no.”
I think you’re dead on about quality being important.
Heck. If I could ditch quantity we could radically slash our infrastructure requirements (but then again we wouldn’t have Spinn3r).
Computing memetracker results on say 50k sites is a LOT easier than 12M.
The point I wanted to make earlier was that the trade off over quantity and quality is a false dichotomy. I think you can do both.
I don’t think we’ve fully proven that out though….. still trying to tune some things.
The Foo,
Clearly you point out some good UI suggestions. Thanks.
I agree with you on a lot of points and this dovetails with some thinking we’re doing for the next release.
Feel free to ping me via email if you’d like beta access before we push it out….
Kevin
“The point I wanted to make earlier was that the trade off over quantity and quality is a false dichotomy. I think you can do both.”
Well, I suggest you start from quality first rather than insisting to do both at same time.
The reason why the news are stale from Mike is because your infrastructure have to crawl millions of site. But frankly,tell me, how much of those information are quality information user care enough to read ?
The problem with your initial approach is that you still try to balance both, and in the end the quantity of information overwhelm the infrastructure ability to deliver quality information.
So I really advise following quality first. You do not need to track millions of site to source for quality news, you just need quality sites to begin with just like techmeme.
I suggest you solve 99% of our user’s need for quality information, rather than remaining 1% for quantity information. Those 1% are not worth solving in term of resource and expense throw at it.
Have you wonder why Techmeme is one man show ? The answer is because Techmeme leverage on few quality sites to deliver quality info to user, and really, that all user really really really care. I do not have much time to read all news, therefore, I want to read quality news.
So Kevin, please take my advice, because no point in believing silver bullet if the gun is rusty to fire anything at all.
Start from quality deliverance first, and then leverage on others resource to do what Techmeme as one person cannot.
Even if you have to start over again, so be it, rather than keep throwing effort and resource at problem that is not a problem in first place.
Good thoughts bigmac. I def agree with a lot of what you’re saying.