Yahoo launched its live streaming service Yahoo Live today, the first major internet company to enter the live streaming space against startups like Justin.tv, Blogtv, Mogulus and others, and the results can be seen in the screenshot above. If we were putting together a report card on Yahoo Live, day one would simply read FAIL.
The idea of Yahoo getting into live streaming isn’t a bad one. The space is a logical vertical for the big players to be in, and there’s been rumors for some time that Google might soon offer a similar service on YouTube. The implementation isn’t rocket science; live video needs bandwidth and solid servers, something the little players are already doing. Yahoo Live isn’t a bad effort, but it lacks some of the obvious value adds of competitors such as the ability to record shows, so in theory it should take less to run…and yet it came crashing down miserably.
Worst still is in the intermittent times it was up, most shows didn’t have huge numbers. I saw one show with 150 people, and Kevin Rose only managed 80 when I tuned in. Consider that top shows on other sites regularly do 400-500 and never skip a beat, so Yahoo’s server loads would be significantly less than competing sites.
If Yahoo cant scale something like this (no matter how much they claim it’s an experiment, it’s still a live service), it shows how far the once brightest star of the online world has fallen, and that perhaps Microsoft could certainly do no worse, and may even help Yahoo. The first thing Microsoft needs to do is send in some smart engineers who understand scaling, and who knows, Yahoo Live Live might have a strong future 
The service looked promising, but it would inappropriate and too soon to jump into conclusions, now.
Duncan: Is an experiment , what part of that word don’t you understand?
techcrunch is becoming more and more “mean” towards some companies. One more useless post
justinkadima
yes, it’s an experiment that fails to scale. It’s a live service on what is still one of the largest internet players out there, scaling should be a given, the experiment side is an excuse, nothing more. Could you see Google launching an “experiment” and it not being able to handle low loads?
Matt
).Yahoo needs saving from itself if this is the standard they set with new product launches.
Sorry that stating facts is mean, how would you sugar coat something like this? (and if I did I’d have a lot more trolls after me than you
hi, duncan, I think you are really picky here, or wait, if you are so picky, why not write a post about the frequent downtime/stability issues of GMail? which has been launched over 4 years now, but still not stable?
Let me get this straight - ever time Yahoo needs some help from now on - TechCrunch is going to suggest M$ sending some guys in? Before Microsoft’s offer they would have just fixed the problem with some bad PR (perhaps). Now it requires Microsoft? Can we all just take a breath?
What a joke. I tried to access this site today and got denied. This is an ok mistake for a cash strapped start-up to make, but Yahoo? Quick, sell to Microsoft.
Yahoo! should launch product like Google or Hulu does. Google in particular, launched GMail through private invitations, so they did not get the surge of traffic immediately once goes live, which is a clever trick to have a good reputation.
I still remember when Yahoo! launched Y! Pipe one year ago, when the news hit slashdot, “the pipe got clogged!” They should have learned this lesson already.
what part of a small “6 member” incubation team you don’t understand?
Don’t judge them too quickly Duncan. Your post is FAIL.
Zeke #6
Gmail is nearly always up when I use it, indeed I cant remember the last time it was down, aside from IMAP support which sucks.
Anonymous #9
Excuses. So it’s a small team, so what, is Yahoo so skinflint that it cant provide a little bit of extra capacity to deal with launch traffic (Employees are irrelevant, were talking about servers and load balancing here). Here’s betting that you’re a Yahoo employee
Laurent #10
Your comment is FAIL my friend. This is a scale issue, can you seriously say that Google wouldn’t be able to handle traffic? If they said it was a coding issue, a tech issue…maybe they need some leeway, but their own error messages said we cant handle the traffic, and it was low traffic at that. Google works in a cloud, Yahoo seemingly works on single servers
Lets them give some more time.
Duncan: Don’t always see the negative side of Yahoo! and compare everything with Google’s services.
Sure this place is getting more biased.
I am an even-hearted supporter of Google and Yahoo!
Zeke, how can you compare gmail, an application with tens of millions of users (lets conservatively say 500K concurrent sessions) to an app that falls over on its first day with a few thousand users? Sure, streaming video takes a lot of horsepower but as Duncan points out, smaller companies are doing a much better job.
http://www.techhui.com
any rumour that Google might set up a live video service too?
That was fast!
No wonder why they are being bought out!
http://www.givemebeats.com
Anonymous
Honestly I’m not anti-Yahoo, or anti-Google, or anti-Microsoft for that matter, but I’ve called all three out previously when they’ve sucked, and in this case it’s Yahoo. If this was a game of winning friends I’ve completely failed, but I’d rather be honest than suck-ass. You can visit some of our competitors for that
Daithi #15
50% maybe more Google will do similar through YouTube, which makes a lot of sense.
> This is a scale issue, can you seriously say that Google
> wouldn’t be able to handle traffic? If they said it was
> a coding issue, a tech issue… maybe they need some leeway,
> but their own error messages said we cant handle the traffic,
> and it was low traffic at that. Google works in a cloud,
> Yahoo seemingly works on single server
Duncan, several of Google’s services have been slow or down during launch. So much in fact that for a while you’d *expected* new services to be down on launch day. In the early weeks of Google Analytics, it even brought down some of my sites with it, as I included their tracking code. Blogger.com’s constant downtimes were part of why I left their service a couple of years ago to switch to, yes, my single server which ran more stable since. Whatever the reason for Yahoo Live’s troubles may be — likely Google allocates a certain *share* of the cloud per service, and perhaps their caching mechanisms need time to roll out, management may scale resources by first checking demand, and what-not, and similar things may be true for other cloud competitors — it’s not completely Yahoo-specific.
The original Google reader was slow as a dog — then about a year after launch they revamped it and scaled it… so this kind of problem isn’t unique to Yahoo. The reason that Microsoft doesn’t have this problem is because no one even cares to try what they launch.
This writeup is total BS. Have you forgotten the numerous times Google put something up live in their “labs” only to have it break down the same day? By that measure, Google has failed at least 15 times and fallen from their run as #1.
Duncan, if you think yahoo, who handles billions of pvs a day, can’t handle scale, you are high. Several experimental launches at yahoo and google have faced early scale issues. You are just picking on Yahoo for no reason here, how far they have fallen, please.
Check out the yahoo home page, it loads in 1.5 seconds. Now look at msn.com: 8.5 seconds. And you think yahoo needs help from microsoft to scale?
Either way its gotta be a PR nightmare for Yahoo
Definitely should have launched a slow private beta first and work out the kinks in my opinion
Smaller companies like Justin.tv ,Ustream and Mogulas have scaling issues with flash crowds when they have popular content on and they seem to be able to respond within minutes .
Yahoo should be able to scale if these little guys can .
Duncan,
They obviously where not expecting the kind of server load that they received after launch.
The real problem with your moronic post is that you suggest that Yahoo! doesn’t know how to scale! (and also in your comments that they do not know how to compute in a cloud).
Basic knowledge of Yahoo!/Engineering understands this:
1) Of course they know how to scale, they keep the largest e-mail and web service running through 100s of millions of users per day.
2) They use a wonderful open source (lilkely with proprietary extensions) tools like Hadoop to run their massive operations.
3) Sometimes things break …
Now, why would Microsoft and Google engineers come in and “save the day”.
Clearly, Yahoo! can do this in house, they have demonstrated clearly through their highly available highly scalable e-mail service that they know how.
Also, Yahoo! owns broadcast.com (an insight I should have gleaned from your superficial post), and has a tremendous amount of in house know how in terms of scaling it.
If you have been paying attention to the internet for more than the “web 2.0″ years, you would also know that Yahoo! has more experience doing video broadcast than any of these upstarts (who simply use a flash streaming server).
So…
You merely should have said.
Yahoo! Your beta was not ready for release. Go back and fix it and come back live when you are ready!
Duncan, I’ve said this before, and I’ll say this again.. Your articles are useless.
Yea BS about covers it -
If the word labs wouldve been embedded in this domain - I am guessing it wouldve passed BS muster
This attitude - internally within Yahoo - and - externally from naive (and muckraking) bloggers like the one above - is what punishes creativity -
First you trash people who have the balls to experiment on this scale - (And they are clearly experimenting here)
And then when they dont because of rubbish from you or their fat ass VPs you trash them again for not doing so
Tough world we live in
Boo. I hope you recant when it’s tuned and you realize what an amazing product it really is.
Arrington, I think it’s time to post an “EDIT”
Also, Yahoo! owns broadcast.com (an insight I should have gleaned from your superficial post), and has a tremendous amount of in house know how in terms of scaling it.
The reason you do “experimental releases” is to find problems. So, they found a problem. Experiment succeeded.
The error message was clearly ill-chosen - something that pointed at the experimental nature of the site and invited people to try again later would have been a lot better - but trashing an experimental release for not scaling is just wrong.
I doubt it’s really a scaling problem - Y! has plenty of experience with scaling…
This is often a big question for companies introducing new services, do they get it out there ASAP to get feedback or spend a lot of time upfront to engineer it to scale. Practically all the web 2.0 startups go with the former option and deal with scalability later on IF they make it big. This makes sense because the feedback can dramatically change a site and hence the scalability needs. However, it probably make sense to initially limit access to the new service to minimize known scalability issues and also create a mystique around it.
Wow… Kevin Rose only had 80 people so there was no traffic? Isn’t it possible that due to Yahoo’s brand and MILLIONS of yahoo users ALL over the world at once tried to create and account/channel and stream their own video…
Comparing to other still small Valley startups with no brand name is kinda weak. When a company like Yahoo launches anything they will beyond doubt see a lot of traffic and the traffic is not users who know who Kevin Rose is… There is a world outside the Valley… NO excuse for not scaling well. But it is pretty brave to open a service like this to everyone.
Btw… I’m pretty sure Yahoo has engineers that can scale!
Is this really news and worth posting about on TechCrunch?
Site launched, saw a lot of traffic, went down. Yes, it’s an error on Yahoo!’s part, but it’s isn’t detrimental in the long-term. It’ll be fixed and soon forgotten and it won’t affect the future of the property.
Please find quality content that keeps me coming back to TC. This post just isn’t it. Focus more on reviewing the service than showing your knowledge of cloud servers.
^++
“Focus more on reviewing the service than showing your knowledge of cloud servers.”
Or complete lack thereof!
Back to CS 101 for you!
I can’t even get the page to load now: “Internet Explorer cannot display the webpage”
I told about 10 people today about this service Yahoo! Who then told 10 people. I applaud Yahoo for stepping up and out.
Live is going to be big!
Top it off with the API coming. It will quickly move from the incubator to one of the main products for Yahoo.
I agree there should be the same or better functionality including recording a show, friends, etc… It will happen.
Right now there are 1,325 people watching 121 live channels. Except to see these numbers double quickly as the word spreads.
@matt (not me): Yeah I’ve noticed that too. Lighten up, folks.
@32 — it’s news because it is Yahoo (they should do better after 14 years) and all the noise surrounding Microhoo.
Gosh, Duncan — I can’t believe you just said you’ve never had issues with google services. Useless post, like someone else already said.
Yahoo is developing a lot of applications that will overthrow google in the near future.
http://www.i-guide.ro
What about “fail early, fail often?” What about “noble failures?” What about the Twice-Shy Entrepreneur? I know this example is not exactly the same as Yahoo Live , but it highlights the ability to push through on an idea and to ignore the fear of failure, or to just not have the fear in the first place. What about learning from failure? Your observation on this failure is very counter to the Silicon Valley view, and this is why SV takes more risks than the rest of the world. In Europe and Asia, failure is bad, in SV failure is a badge of honor.
I’m pretty sure your posting unfair articles like this just to spark debate so your flooded with traffic and comments. There I go posting a comment on the article! How ironic.
This wasn’t an official launch. Do you really expect Yahoo to invest in large amounts of expensive servers just for an experimental release? That would really please the stockholders… “We don’t know how many users will come, so we just buy a hundred servers just to be sure…”