on Jan 16, 2012
Last week I mentioned that I was changing advertising providers. I tried Project Wonderful, and ran one of their ads on top of my site for several days. I don’t expect most of you to care, but for those of you considering running ads on your own site or who are curious about how this all works, here is what I’ve learned:
I am not an expert and most of this analysis is conjecture based on my own experience as a purveyor of entertainment on a medium-sized website.
For years I’ve had a single Google “skyscraper” ad running in the sidebar. Google ads are very hands-off for the publisher. You don’t have a way to see what ads are running. If people complain about an ad (“since when does your blog support obvious scams, Shamus?!?”) you have no way of perusing the active ads to find what company is causing the problem. If you’re lucky enough to see the ad yourself, you can only block it by clicking on the ad itself, (which is technically a TOS violation for me to do, but hell if I’m not going to protect my site) seeing what domain it goes to, and then navigating around in the Google adsense interface and copy & paste the offending domain into the clunky block list. Then several hours later that ad will stop showing up. I can’t block by company, which means domain-hoppers need to be blocked again and again.
Ads can be region-locked, so there are often ads running that I will never see. I get angry emails from people asking me how I could, as a [Christian / programmer / father of daughters / decent person] ever allow such [smut / scams / misogyny / trash] on my website. They don’t realize that I’m not seeing the same thing they are. People naturally hold me accountable for what they see on my site, which wouldn’t be a problem for me if I had the proper tools to know what was appearing on my site.
Google Adsense makes it impossible for me to see everything appearing on my site, and hard for me to police. Google says that advertisers prefer prominent ad space, but there aren’t any discernible incentives for me to give them that space. As far as I can tell, I get paid the same for ads on top as ads on the bottom.
On project Wonderful, your site enters a gigantic, organic, ongoing, chaotic auction where advertisers pay for time on your site. It works like this:
Aardvark Games decides they like the look of your site, and decides to advertise on it. They bid to pay $1 a day to have their ad on your site. Because they are the ONLY bidder, they don’t actually have to pay anything yet. Their ad runs for free.
Bison Publishing comes along and decides they want in. They offer a nickel a day to be on your site. Aardvark Games now automatically moves to beating this bid by one penny. So now ads running on your site make you six cents a day. $0.06.
Caribou Marketing joins in. They offer fifty cents. Again, the Aardvark bid moves to beating the second-highest bidder by a penny. Your site is now making fifty-one cents. $0.51.
Donkey Entertainment offers $2 for your site. Aardvark is now the #2 bidder, so Donkey ads replace the Aardvark ads and Donkey begins paying you one dollar and one cent. $1.01.
So at any given time, your site is worth as much as the second-highest bid plus one cent.
You can always get a listing of who has open bids on your site, how much they’re paying, and how long the offer will last. You can view the ad. You can visit the site being advertised. You can block it. You can set filters for your site that range from “safe for kids” to “safe for work” to “anything goes”. I can decide if I want animated ads or not. This is all handled in an interface that’s actually kind of fun to use.
You can set a minimum bid on your site, or you can let the market have its way with you. People bidding can see your site traffic and get a sense of where their ad will appear before they bid.
This is, like it says on the tin, a wonderful system. It’s friendly to use, it’s transparent about where ads are coming from, and it gives you a robust toolbox for dealing with people who abuse your hospitality.
Project Wonderful would be the better service by far, if not for the fact that you’ll make about an order of magnitude less money. Actually, even a single order of magnitude is an optimistic projection. My site brings in not less than $4 a day through Google. That’s not enough to feed a family or anything crazy like that, but that’s over a hundred bucks a month, minimum, for a site my size. Through Project Wonderful, I have yet to earn even twenty cents a day. That works out to six dollars a month.
Now, maybe my income would go up if I stuck around in Project Wonderful and became a popular site among advertisers. However, I’ve searched through the auctions, and even the top-level earners (among similar sized sites) don’t make anywhere near what I’m bringing in with a single Google ad. And that’s for the special few at the very top of the curve. (Which is another strange thing about Project Wonderful. I can search for sites and see exactly how much they make per day. Was kind of interesting to see how much something like Girl Genius brings in. Short answer: WAY more than me, but probably still not enough to live on.) The vast majority of sites in my range bring in just over a dime a day for one ad. So, 1/40th of what I make through Google?
Because of this, it’s generally not worth it for a lot of publishers to bother claiming their six bucks at the end of the month. Instead they seem to turn around and sink this cash into advertising their own site. (A lot of these sites are webcomics.) So, maybe webcomic X is making $0.12 a day running ads and then paying a couple of smaller sites $0.06 a day to run an ad for Webcomic X. What we end up with is a convoluted link exchange.
This is not actually the fault of Project Wonderful. They have a brilliant system, and it seems like this is how advertising should work. The only problem is that big advertisers aren’t interested.
I’ve never been an industry insider and I’m not privy to the direct deals that are brokered between your average gaming site and your typical publisher. By a very rough but very conservative guess, I’d say that a big gaming site can make about ten times as much as I do for the same traffic. Of course, it takes them half an hour to hit the traffic that I’ll do in a day, but on a cost-per-exposure basis, they make at least ten times what I do. (And really, those sites support an entire STAFF of writers, so I’m sure my figure is on the far low side.)
So the big dogs like IGN make ten times as much as I do for every pageview. And I make forty times as much per pageview as your average Project Wonderful site. Which means advertising on Project Wonderful is (conservatively) 400 times cheaper than advertising on (say) Gamespot. For the cost of one banner at at the top of a big site for one day, you could occupy the banner space of 1,000 different blogs and webcomics. For weeks. To my knowledge, this has never even been attempted. I suddenly feel a lot less sorry for publishers and their complaints about how expensive advertising is.
I think part of the problem is the Project Wonderful model of paying for days. Advertisers care about exposure. They care about having their rectangle.jpg files in front of as many eyeballs as possible. This model is like going to the deli and paying for meat by how much is cut in ten seconds instead of going by weight. It’s an abstraction for advertisers, and puts more work on them. It also makes the exchange a lot less certain. You could bid on my site on a Friday, seeing that the metrics are saying I’ve averaged 35k pageviews a day for the last five days. Then your bid goes through on Saturday when I never post and my traffic drops by 40%. Or maybe I get a spike of traffic from Reddit that inflates my averages for the next five days. Suddenly you feel ripped off, because you paid for a day on my site and got a fraction of the exposure you expected..
It’s not reasonable to expect an advertiser to be able to follow the traffic patterns of all of the dozens or hundreds of websites they might advertise on. It might be cheaper this way, but it’s not how the industry has evolved and it’s not how people are used to buying advertising.
I see where the Project Wonderful mindset comes from. A lot of scummy sites really poisoned the well in the late 90’s and early 2000’s. There were auto-refreshing sites and scams that goaded visitors into clicking on ads they didn’t care about. There were sites that promised content (I vividly remember being a big user of no-CD cracks when I was in my twenties, because I had a horrible habit of leaving discs on my desk until they got scratched) but were really just traps to lead frustrated visitors through circular links that showed wall after wall of banner ads. A small number of creeps polluted the entire business model for both parties, for years. Advertisers flocked to big, reputable brokers like Google, and websurfers invented adblock. It fostered an adversarial relationship between producer and prospective customer.
In world with less jerks… well, a lot of things would be better, including this.
The upshot is that producers don’t want to advertise on small sites, because they can’t babysit a thousand small sites to make sure none of them have “PLEASE CLICK ON MY ADS” in big flashing letters.
Project Wonderful sidesteps this problem by having you pay for time on a site instead of clicks or pageviews. Now there’s no longer an incentive for publishers like me to goad you into refreshing the page or clicking on the ads. The stuff is just there, and will garner your attention based on its own merits.
It’s an interesting idea, but the results speak for themselves: Despite the rock-bottom pricing, there still aren’t many interested advertisers. You could argue that advertisers are dumb to pay 20 or even 400 times as much for the same exposure elsewhere. You could be right. But I wouldn’t hold my breath waiting for an industry this large to change direction.
I’d love to see a system exactly like Project Wonderful, but based on the price of pageviews. Maybe advertisers would show up because it’s a pricing model they understand, or maybe the well would again be poisoned by people trying to game the system.
A lot of this is a problem because the system is so asymmetrical. On TV, a big advertiser comes to a big TV station. Both parties are large enough to have lawyers and contracts and a lot of long-term incentives to avoid cheating. When the campaign is over, there’s no ambiguity over how much advertising was done because it was all done in public where anyone could measure it. On the web, you have a big producers and tens of thousands of bloggers. There are no lawyers, and the “contract” basically boils down to a terms of service agreement. The bloggers have little accountability and the advertisers don’t have a good way to keep tabs on everyone.
It’s an interesting problem.
I liked Project Wonderful, but at this point I’m going to take my accumulated $0.53 and buy myself a pack of gum. The PW ads are gone. I hope some advertisers show up soon and give the rest of these sites some much-deserved love. And by love I mean cash.
Shamus Young is an old-school OpenGL programmer, author, and composer. He runs this site and if anything is broken you should probably blame him.