Stack Overflow just became a top-500 site. | Hacker News Reader

Stack Overflow just became a top-500 site.

quantcast.com

104 pointsdb4251 comments

It's almost hard to believe we went so long without a site like StackOverflow. Not surprising they've grown so fast considering how much better they are than everyone else.

Congrats & thank you to Jeff & Joel & the rest of the SO team. You guys truly make the Internet a better place for all programmers!

Yet ExpertSexChange is still near the top of Google search results - 'tis a mystery. Still, it validates the SEO business model that gaming the algorithm can get you higher ranked than real actual content of the kind SO has in buckets.

Experts exchange has been in business since at least 2001, they also have a bit actual content.

I think you can still just read Google's cached version for those results. The actual content is below the placeholder content.

Ugh, I wish people would stop thinking it's clever to call Experts Exchange "ExpertSexChange". First, it's not clever, and second, it implies that there's something wrong with a sex change.

Ah yes, dismissive argument format #2837: "I hate when people say $foo. It's like $politically_incorrect_thing."

The problem is that calling "experts exchange" "expert sex change" does not say anything about the validity of a sex change operation. It's saying that it's a dumb name for a programming Q&A site. This makes it equivalent to saying, "I hate ExpertsExchange. Look at how dumb the name is."

I think you're wrong there Jonathan. The argument is that calling it "expert sex change" is like deriding eBay as "eGay" (I know I had to change a letter there: it's similar enough to be the same kind of thing.)

It's misusing something important and personal as a simple insult. It's impolite to use something like "sex change" as an insult. It's not "political correctness gone mad," it's good manners, and I think using it shows the author is either thoughtless or bigoted.

It is also a weak, ridiculous and puerile insult that adds no value to any debate. If Experts Exchange was awesome, we wouldn't criticise it for its name. We'd probably not even have noticed.

any experts exchange result has the actual responses down at the bottom of the screen when you scroll down. They aren't "gaming" anything, the content is plainly there, just far below the fold.

Your comment validates my belief that most people don't understand SEO and think it's all evil

By ensuring their landing pages are as unusable as possible Experts Exchange is reducing the usefulness of search results. They are gaming the system to the detriment of everyone else and this is a perfect example of "evil" SEO.

The only reason this is the case is Google threatened to de-list them unless they showed the actual content the page was being cached for. They didn't tell anyone they shifted it down the bottom though, and the only way you can see that content is by using a search engine as a referring URL.

Not all SEO is evil but that one is a bit naughty.

Edit: spelling is awesome

There was a time when just above the answer was a license saying you weren't allowed to read the answer — but it's gone now, so I guess Google made then remove it. I didn't realise on my own that you could scroll down to the answers though, as the page is designed to make that appear not to be the case, so I'd say it is evil SEO.

I really love SO. The thing that really got me started using it was the article I read (on here) about a guy who got headhunted for a job interview at Google off the back of his stack overflow score.

I thought - holy geez, I'd better get working then!

I've not actually been to quantcast before but it seems an interesting site. I usually grab my rankings data from Compete and Alexa

Top 300 sites worldwide and US in Alexa http://www.alexa.com/siteinfo/stackoverflow.com# Top 11,000 in Compete (maybe it's a bit better if you have their PRO ranking) http://siteanalytics.compete.com/stackoverflow.com/

Actually... Quantcast looks like it might be fundamentally different from Alexa and Compete. I wonder how much it costs? Maybe I could use that to find my audience.

One thing I found interestedin on quantcast the location dropdown and see how the site ranks in other countries.

Out of interest, I pulled Canada and saw that we've got Stack ranked as 105 instead of 500 ... which must mean we're 5 times "techie-er" than the States. BOO YAH! Then I switched to pulling the other dropdowns and realized we're both getting our butts handed to us by the UK, Germany and India. Probably just means all their OTHER sites suck eggs. :)

quantcast (we) are free

http://www.quantcast.com/info/measure

Heh that's pretty funny as I believe I'm that guy.

As a bit of a teaser I have a followup coming to that particular post.

Hint: I'm moving to New York. :)

I've never looked at quantcast stats before. How accurate are their estimates of age, income, kids? I ask because it says that 23% of Stack Overflow's visitors are female, and that doesn't seem very accurate. NB: This is based solely on my recollection posters' avatars, which may be a bad measure, but it's the only one I've got.

In my experience, women on the internet tend to be less likely to indicate that in their profiles/avatars.

About a third of IT staff at my employer are female. That seems reasonable to me.

It would be interesting to find out how many users they're getting from their semi recently launched API (http://stackapps.com). I know my iPhone Stack Exchange client http://sixtoeightapp.com has over a thousand downloads, and won't be counted by Quantcast. Whilst it's more of a complementary access point, other apps on Stack Apps might suffice as the sole access point for passive users (i.e. those who just want answers, as the API is read only currently).

Nice one, checking your app out now...

Offtopic - how on Earth did Apple give you a 12+ rating for an app about programming questions!? Another mystery.

I chose it. Any app that downloads data from the web needs to be careful with age ratings - Stack Exchange has more than programming questions - it has, e.g. video games, with perhaps more age restricted material. I chose 12+ as a compromise between Apple's desire to have age approprate labelling, and restricting likely users.

Nice app! Any plans for an iPad version?

Yup! Not sure exactly when, but certainly on the radar.

In under two years, no less! It shows what dedication, attention to the big details, and a clear _goal_ can get you.

"So long hyphen site."

Think about the demographics of people who seek out and install the Quantcast/Compete/Alexa toolbar. Now look at the demographics of their respective Top N lists.

Now imagine a knitting community site that gets twice the traffic of StackOverflow. Would it appear in the Quantcast/Compete/Alexa Top 500?

StackOverflow is a Quantcast Quantified site. This means it's traffic is directly measured with a tracking pixel on each of Stackoverflow's pages.

The problem is: Are all 499 sites above them measured the same way? And all the ones below?

The answer is no, of course. Most sites don't run the QC tracking pixel. That makes any ranking they give fairly questionable, in my mind.

case in point, we (http://minecraftforum.net & http://minecraftwiki.net) have 60m page views per month (as Stackoverflow does -- according to Joel) and we're nowhere near the top 500 according to QC.

The ranking is based on visitors, not page views. Based on the graphs posted a few weeks ago at http://imgur.com/a/haHHz, it looks like you have about 2.8 million unique visitors per month to SO's 8.6 million in Quantcast

This is uniques? Ah, that changes things then ha. Although none of those graphs there list our uniques (I don't think? Don't match the figures I have here and GA calls them "Uniques" explicitly, they're visitors in those graphs) we are definitely lower than the 8m SO has.

I checked, the graphs you linked are visits. As of today our visits is at just shy of 11,000,000.

Edit: Ah, you're right. I see, under "visitors". Yeah, that looks about right, we have closer to 3m uniques. So many different terms in stats :D I wonder what the value of uniques vs. page views is.

Uniques are what count for big ad campaigns. Well, mostly. Uniques get the campaigns which you then have to fulfill with pageviews.

Ravelry.com?

I guess that settles the debate about whether choice of language matters or not :)

Although I can't imagine scaling has been fun (or cheap).

Jeff wrote about PHP, the language does not matter: http://www.codinghorror.com/blog/2008/05/php-sucks-but-it-do...

I've never seen language choice as a big deal. It's just a tool among many others to get a product done.

From what I've read on their blog they seem to get along fine with a moderately expensive SQL server and some chap application servers hosted in a data center. No fancy NOSQL stuff or cloud instances involved.

Scaling has been cheap, and fun.

Read about our newest data center: http://blog.serverfault.com/post/1432571770/stack-overflows-...

Read about our SQL Server hardware: http://blog.stackoverflow.com/2010/10/database-upgrade/

Are you factoring in the SQL Server licenses when you say it is cheap?

Perhaps they're taking advantage of BizSpark or WebSpark or a similar programs. If they aren't, the dual CPU machine they mention would require $14,000 of SQL Server licenses (if you paid list).

EDIT: Actually there's a newer "Web Edition" targeted to roles like this that halves the price. So $7,000 for that server (the details are sparse, but it also sounds like you can opt into it for $15 / month / processor).

That's a lot to someone trying to shoehorn a startup, but is it really a lot for real businesses?

On the server end they still have tremendous headroom in the scaling up department...and that's before the Westmere-EX drops and bumps the top up quite a significant amount. You can add 1TB of memory, and four 8-core processors, to commodity Dell servers. That's before you get into the serious business.

Bizspark is a short-term program, where you invalidate yourself when you get revenue. It's a good idea but I understand it only delays those looming SQL Server per proc license costs.

While I congratulate Jeff and Joel I would like to see more numbers about the 'cheap and easy' claim if possible please?

I don't think anyone pays list price for SQL server licenses. They're still not cheap, but the suggested prices are usually 10 to 20 percent high.

The technology stack of StackOverflow is excellent. The scalability potential is rationally limitless.

Joel, Jeff and crew did a great job. Huge kudos to them.

Grooveshark is bigger than StackOverflow, and our page views/server ratio is better too. And we run PHP. The language argument is BS. I will say that Microsoft SQL server is easier to scale than MySQL. But the licenses are significantly more expensive. :)

You're also much more immoral :-) Profiting off of music without licenses that grant you permission, a good business there.

I'd be happy to discuss the morality/legality of Grooveshark in another thread. It's not relevant to this discussion at all.

Stackoverflow is an excellent medium for asking programming questions.

The thing I question is their VC-backed strategy to apply that to many other fields. Sure it's early days yet but even the original "trilogy" sites haven't come anywhere near SO's success.

I have this nasty feeling that SO won't be able to justify that investment and thus turn from a successful small business to a failed venture.

On the plus side only one USV portfolio company (I forget the name) has ever switched off the lights.

Is this directly due to them being featured on MSDN?

It seems that they have stable growth and there is no big increase in traffic last time. So it looks unrelated.

I wonder what the numbers are for all the new sites, and whether they're started gaining serious traction outside of the SO regulars.

Alexa has had them in the top 500 for a while:

http://www.alexa.com/siteinfo/stackoverflow.com

That's very impressive, I think that stackoverflow will see a major boost from google groups shutting down too.

Google Group isn't shutting down. They're turning off files and pages, that's all.

right:

http://groups-announcements.blogspot.com/2010/09/notice-abou...

Reply on news.ycombinator.com