Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How We Make Money at Stack Overflow

stackoverflow.blog

Post reply on HN
Print view View original

181–190 of 227 posts

Re: How We Make Money at Stack Overflow

#181

Post by jazoom »

secretRubyDev wrote: post #113
tgb wrote: post #97

Earlier quoted context omitted.

Why do they do that? Seems great for this use-case but terrible information to give bad actors.

When I've usually clicked the report spam and unsubscribe button after reporting spam in gmail it usually immediately opens a tab with the unsubscribe URL. I've since stopped doing it since the last thing I want to do is open up some random page belonging to the person who spammed me.

That's strange. I thought the point of that button was to overcome the exact problem you just described.

Re: How We Make Money at Stack Overflow

#182

Post by pcurve »

kevin_thibedeau wrote: post #177

Seeing their salary calculator makes me depressed what a 1-year experience 1x developer in the NYC metro makes.

Consider how depressed developer salary is in Japan, I'm surprised how well Stackover pays... I'm guessing their Japanese office is tiny.

Re: How We Make Money at Stack Overflow

#183

Post by brazzledazzle »

angry-hacker wrote: post #163
techdmn wrote: post #158

Earlier quoted context omitted.

It confirms that the address is valid, and most likely attended by a real person.

But there are gazillion email leaks. Go scrape them all with hashed passwords. Active emails have no value.

You're right that they don't have value but there's still plenty of idiots out there buying and selling useless lists and while they're worthless they have a cost to us as recipients since we have to mark them as spam and delete them.

Re: How We Make Money at Stack Overflow

#184

Post by jack9 »

ocoster wrote: post #28
jack9 wrote: post #19

> And we don’t want to use an automated system that selects some ads for us. So they do direct deals mostly, which is antiquated and probably to preserve their branding as a "tech site". > Did you know we have a lot of unpaid inventory on Stack Overflow every month? This is what trying to guess your avails with direct ends up doing. Leaving inventory unmonetized. Typically, you get paid less for less of your inventor…

What you are missing is that there is a cost associated with monetizing the full inventory with automation. You lose control over what is being shown, you are diluting your brand and you are at risk of alienating your users. At Stack Overflow we are not willing to make this trade off.

> What you are missing

Automation costs are actually administrative costs (platform fees), which have recently increased via unmonetized opportunity cost models (opportunity being a specific technical term in advertising). SO is not fully or optimally monetizing, probably because they prefer to keep the old tech they are using and/or the staff that has been there since early days. It's also possible they want some dejour/nonstandard controls. Saying they are making an informed tradeoff is simply incorrect. Saying they are making a (unspecified) choice, then not really talking about that choice, is what the article skirts.

Sub 10k/mo for platforms at 2 billion opps per day is AOL and SpotX. That schedule doesn't matter at the volume SO works with. Fees and Opp costs are not the issue with the unmonetized inventory.

Re: How We Make Money at Stack Overflow

#185

Post by smsm42 »

angry-hacker wrote: post #163
techdmn wrote: post #158

Earlier quoted context omitted.

It confirms that the address is valid, and most likely attended by a real person.

But there are gazillion email leaks. Go scrape them all with hashed passwords. Active emails have no value.

Leaks may have emails that are defunct, belong to dormant accounts nobody reads, throwaway accounts like mailinator.com, honeypots, etc. This confirms the account belongs to a real human actively reading the email.

Re: How We Make Money at Stack Overflow

#187

Post by morgante »

pbhjpbhj wrote: post #170
nicky0 wrote: post #45

Earlier quoted context omitted.

Most of the ads are served to non-logged-in visitors who come to specific question pages through a web search. SO actually shows you fewer ads the more reputation points you have.

What's the flow there for a user? They're researching an answer ... do they ignore the issue and run off and buy something? I sometime have poor focus but when you're problem solving it seems the least likely time to distract yourself with shopping. I did once follow a Google ad about 10 years ago, but in my defence I was just surfing cool tech and the image looked interesting. Of course I didn't buy anything, who le…

> They're researching an answer ... do they ignore the issue and run off and buy something?

Yes, that's an entirely credible idea. To use the first ad I found as an example, maybe I'm debugging an annoying issue with migrating my MongoDB cluster and I notice an ad for a managed MongoDB service and realize I'd rather pay them to deal with this.

> Of course I didn't buy anything, who lets advertisers tell them what to buy?

I do! There are millions of useful services out there and ads are an effective discovery tool.

Re: How We Make Money at Stack Overflow

#188

Post by morgante »

stevvve wrote: post #146
huhtenberg wrote: post #73

Earlier quoted context omitted.

Re: balancing the needs of advertisers - We looked at advertising with them about a year ago and their options and response were both underwhelming. It wasn't possible to geo-target, it wasn't possible to advertise on most relevant sites (e.g. SoftwareRecs), and their upfront budget requirements weren't compatible with how we normally do things, which is to run a small test campaign to see what sort of traffic an out…

Hi there! I'm Steve Feldman and I manage the Ad Ops team at Stack Overflow. I'd like to make a few leaps based on your comment and offer a little insight our business. I'm very interested in learning more about your experience. - Based on your targeting concerns where we failed to adequately meet your request (geo, site, budget), I'm going to assume you tried to purchase advertising through our self-serve option. We…

I've managed larger advertising campaigns which would easily meet your $5k/m minimum but agree with the GP that your ad structure is a non-starter for me.

Whenever I've looked at introducing a new advertising channel, I always start with a test run of a few hundred dollars to see if it's worth investing more. I'm not going to drop $5k on an ad without knowing if I'll get any results at all.

It seems like you only offer two options for someone like me:

1. Run a test campaign without targeting and therefore get suboptimal results. This likely just lost you the sale of a bigger campaign (where I might get good results with targeting).

2. Risk spending $5,000 on a channel which doesn't give effective results. This isn't a risk I'd usually be willing to take.

Maybe you're targeting brands who are big enough to experiment with $5k buys or who don't test channels. Either way, I'm really not sure what the value prop is on the self-serve option.

Re: How We Make Money at Stack Overflow

#189

Post by morgante »

ComteDeLaFere wrote: post #78
antaviana wrote: post #39

Earlier quoted context omitted.

According to Hoovers, their revenue is $11.33M. Source http://www.hoovers.com/company-information/company-search.ht...

How do they employ almost 300 people on $11 mil / year? Also, why are these posts getting downvoted? The title of this article is "How We Make Money at Stack Overflow".

Venture capital. I highly doubt they're even close to profitable with revenue per employee of $36k.

Re: How We Make Money at Stack Overflow

#190

Post by Shog9 »

0xmohit wrote: post #104
finnh wrote: post #52

Earlier quoted context omitted.

How many questions does it take before you are banned?

If they started banning users for asking questions or even for asking poor (read extremely poor) questions, then the overall traffic would reduce significantly .

I wish. We've been doing this for 5-6 years now, and it hasn't. Quite honestly, if we could get rid of the worst 20% of questions instantly, life would be a lot better for the remaining 80%... But, identifying those 20% when often SO is their only option is... Brutally hard. I've seen the same person try to post variations on the same question dozens of times before finally getting around the quality checks... Only to have it sink like a rock because of course it was still unintelligible after all that.
Post reply on HN