OnStartups

Online Advertising Revenue Buys Me A Chai Latte

Posted by Dharmesh Shah on June 11, 2006 20 Comments

Those of you that are regular readers of OnStartups.com may recall that I had turned on online ads on the site for about a week.  As I noted back then, the motivation for this was more educational than financial (i.e. I wasn’t expecting to make much, and that was not the primary motivation).  Ads were left on for a week and then turned back off again.
This time around, I tested with Yahoo Publisher’s Network (YPN) – though I’ve run a similar experiment in the past with Google AdSense on another site.  I saw better results with AdSense overall, but not enough to change my opinion in any meaningful way.
In the week that I had advertising on, I generated just enough revenue to buy a nice cup of coffee at Starbucks.  No, I don’t mean per day, I mean revenue for the whole week.  The week that I ran the experiment was a relatively typical in terms of traffic on the site (clocking on average about 1,500 unique visitors a day).
Here are the likely arguments the online advertising experts would make regarding why the revenue generation was so low.  
  1. AdSense would have generated more:  I agree, it probably would have and I’ve tried that before too, but it would not have been more than 2 times more than YPN.
  1. Ads were not well placed:  I agree.  I was not looking to make ads the focus of the site in order to maximize revenue.  I wanted to see what would happen when I just placed them where it felt “right”.
  1. Not the right audience:  I agree.  OnStartups.com is not about gadgets, luxury travel or other “transactional” content that would likely be of interest to high-end advertisers.  It’s for entrepreneurs that are likely more interested in content and are not really looking to make a specific type of purchase.
  1. There are other advertising vehicles:  I agree.  I could have gone with CPM models, instead of CPC.  I could have used Chitika (which I have also tried before), or found more targeted advertisers myself.

  1. I don’t know what I’m doing.  I agree.  I’m by no means anywhere close to being an expert in online advertising and I’m sure if I had more knowledge and experience, I could have generated more revenue.

As noted, I agree with all of the above points.  This experiment was by no means conclusive nor scientific in any shape or form.
The big message I have here is this:  Advertising revenue is not easy.  You can’t just plop an ad up on a site and hope to watch the money roll in.  Though I could likely have increased my success-rate (possibly by an order of magnitude), this requires some degree of insight, learning and effort.  Personally, this is not something I’m particularly interested in doing.  For every hour that I could spend learning about the best way to generate ad dollars, I’d prefer devoting that time to reading more about my area of interest and writing content that I think would be useful to you, my reader.  Even if I was able to increase my ad revenue 100 fold, I’d still be making about $500/week, which is simply not interesting enough to be worth the effort.  And, it would be a lot of effort.
I would make a similar argument for the software startups that are hoping to use advertising revenue as their primary revenue vehicle.  In order to really do this, you have to:
a)     Have lots and lots of web traffic.  This is not easy to do, even for a relatively successful site.
b)     Ultimately be able to convert the traffic into meaningful interactions that advertisers are willing to pay for.
c)      Spend some amount of energy learning and experimenting with how to optimize the advertising revenue generated.
Though this is certainly not impossible (as proven by a number of exceptional examples making lots of money in online advertising), I would argue that most software startups are not necessarily good at knowing how to make meaningful money with advertising.  Few software developers are born with the skills required to be great online advertising revenue generators.  Though it’s learnable, you have to make a conscious choice as to whether you want to spend time learning about creating better products – or creating better placed ads with the right level of intrusion to optimize the money you can make.
Further, if I were putting ads on a software site (like my current startup, HubSpot), it would be even harder for Yahoo! or Google to generate meaningful ads.  Most of the content they would need to select the right kind of ads would be hidden behind a secure sign-on.  Most software startups, other than those focused on user-generated and publicly facing content would manifest this challenge.   Without a lot of meaningful content, the advertising engines can’t place the right ads.  This is why I do not expect to use advertising as a means to generate revenue for my startup.  It just doesn’t make sense for me.

Summary Of My Point: 
Software startups relying primarily on advertising revenue need to experiment and actually try it to see what kind of money they can currently make.  For example, if a site is getting 100,000 unique visitors today and making $100/week in ads, we can extrapolate a little bit from there and figure out what kind of traffic would have to be generated to make enough money to be worthwhile.  Chances are, it’s going to be a lot of traffic.