We're building our own ad network | Hacker News Reader

We're building our own ad network

stormpulse.wordpress.com

22 pointswensing9 comments

I wish them luck, but I don't think selling ads on a weather site will be easy, especially without a massive amount of traffic. What types of ads will be relevant? Those for umbrellas? Maybe products for storm chasers or meteorologists? These types of businesses are probably not in the habit of buying internet ads, and if they give it a shot, it likely won't convert.

Looking at other weather websites, the ads are generic and untargeted. Basically massive brand campaigns for prescription drugs, insurance, etc. Weather is likely a poor monetizing niche and these advertisers buy massive amounts of remnant inventory.

General weather forecasts may be a poor monetizing niche (because 5-day forecasts are a commodity), but severe weather is completely different. Depending on where you've grown up, this may be more or less obvious.

Niche-relevant ad networks are definitely going to be in demand these days...

If anyone knows one which primarily reaches educators, I'll be over that like white on rice.

Here are my issues with the non-Google online advertising I've tried so far, and I've tried a bunch of them.

1) I sell a really niche product. You deliver a really niche audience. It is possible that my niche times your niche equals a number of interested prospects so low that it is not worth my (very expensive) time to investigate, implement, and optimize your ads. Remember, advertising on a competing website who uses AdWords costs me no time at all at the margin.

2) I only want to pay for traffic that has a chance of converting. One provider I used, who will remain nameless, charge me the same prices Google does. However, my ad (for a product which is culturally only well-suited to American customers) was getting displayed in the middle of education websites in Spain. And I was getting charged full price for that.

3) My time is very, very expensive -- I shoot for $100 an hour. Let's say it takes, oh, two hours to get everything set up for your website, including conversion tracking, writing copy to conform to your requirements, etc. And lets say that you deliver sales to me at, oh, $10 apiece.

$30 product - $10 for you = $20 for me, per sale. (We'll ignore my other marginal costs since I essentially don't have any.) $200 / $20 = 10 sales required to break even against my time. 2% conversion rate implies I usually need roughly 500 free trial downloads to generate 10 sales. 20% conversion to free trial implies I would need 2,500 clicks from interested users to generate 500 free downloads.

So to just break even with my time for setting up your service once, I need 2,500 clicks from interested users. That is if you deliver absurdly high ROI immediately right out of the gate. If you don't, then I need to analyze results and optimize my ads for you, and that again has an imputed value of $100 per hour.

Most non-Google advertising platforms I have tried require far too much of my time to justify the returns they have provided.

"3) My time is very, very expensive -- I shoot for $100 an hour." - Aren't sales the life blood of your business? Is there anything else more important? Do other sales channels cost you less time and produce a higher return on investment? It almost sounds like you're looking for the goose that lays golden eggs. It's like you want to make a sale without any effort at all or if you do put time/effort in - you want paid "extra" for your time - doesn't the sale of product pay your salary?

p.s. I have no affiliation with stormpluse - although I love their site.

I agree that sales are important, which is why everything I do should generate them. SEO, tweaking website copy, investigating advertising channels, program updates, etc.

I've got a million things to do and a fixed number of hours to do them in. I prioritize based on expected marginal sales generated by marginal effort. Anything which can't hit $100 in marginal sales per marginal hour gets eliminated, automated, our outsourced. (Unless I really like doing it -- however, running advertising campaigns is definitely on my list of "Shoot, I would do this for free" activities.)

In my experience, methods of advertising to my niche other than AdWords fail to generate sufficient sales per increment of time to justify using them when I could be doing much, much higher ROI activities. AdWords has insanely efficient ROI on time for me these days. I log in once a month for about twenty minutes and note "Oh, $600 in ads produced $1,200 in sales -- well, carry on then." and then pencil it in for another checkup in a month.

What technology do you want to use? I think that you can use openX or Adify.

We're looking at using Google Ad Manager. https://www.google.com/admanager/

Anyone have any experience with it?

It depends on what they plan on doing - if they are running ads just on Stormpulse then there are lots of solutions available.

If they are building a vertical ad network (i.e. connecting advertisers with multiple weather-related publishers) than something like Adify will do the trick.

Full disclosure: I work at Adify.

Reply on news.ycombinator.com