Earlier this week I mentioned that I had started using Conversion Optimizer, Google’s new thingamagig where you bid Cost-Per-Action (CPA) on the content network instead of Cost-Per-Click (CPC) and let their black magic sort it out.  The preliminary results are in: this is the best thing to happen to AdWords since AdWords.  Not only is it a totally automated smackdown of the most egregious spam sites (and quite similar to a solution I proposed a few days ago), it has drastically cut my advertising expenses while not hurting revenue that much at all.

Pre-Optimizer Daily Spend: $10-12

Post-Optimizer Daily Spend: $5-6 (down 50%)

Pre-O CTR: .8 to .95 %

Post-O CTR: 1.1 – 1.2 %  (up 50%)

Pre-O Clicks Per Day: 100 to 120

Post-O Clicks Per Day: 60 to 80

Pre-O CPA: 45 cents to 50 cents

Post-O CPA: 25 cents (down 50%+)

Twenty five cents.  Twenty five cents.  Wow, has it been a while since I’ve seen that number. 

Lets do the math here, shall we?  Assume that all trial users are equally likely to convert, which I cannot substantiate but believe to be more or less accurate when we’re talking in aggregate.  My conversion from a trial to purchase is, in round numbers, 2.5%.  If I pay 45 cents for a trial, then I am paying $18 to make a sale worth approximately $24 to me.  I still make money, but Google and the content publisher get most of the value from the sale, and I do all of the work.  Boo that!  And if I’m paying 50 cents for a trial, then I pay $20 for a sale worth $24, and at that price I might as well just not run the ad at all. 

If, on the other hand, I get trials at 25 cents, then one purchase at the margin costs $10.  Much, much better.  That doubles my profit margin without me lifting a finger, other than to turn Conversion Optimizer on.  Google and the content provider also are getting decent returns, which ensures that both will want to continue doing business with me.  For once the interests of Google, the content provider, and the advertiser are in perfect alignment here: I get the best CPA on sites which are closest to my niche, which means the readership is more motivated, which raises their CTR on my ads.  (This is because visitors at Mrs. Lindle’s Reading Tips are much more likely to want to play bingo today than folks reading snazzygurl413’s MySpace page.)  A higher CTR means that Google gets more for the same amount of ad inventory.  It also means that Mrs. Lindle gets more money, because they are putting high performing Bingo Card Creator ads on her site instead of the random garbage that Google usually fills low-CPC niches with, and even MySpace benefits, because they are no longer “selling” me 10,000 impressions a week which never generate so much as a single 8 cent click. 

When Google does things right, wow, do they do things right.

What is the downside?  Well, if you’re the tinfoil hat sort, you can worry about this one: to do this, Google needs to know exactly how much a conversion is worth to me.  That would, theoretically, allow them to achieve the holy grail of monopolies: perfect price discrimination.  That describes a situation where the consumer surplus (extra value you capture because you pay less for a good than what you value it at) vanishes, and this maximizes the firm’s potential profits.  Joel Spolsky mentioned that the market effect of two competing firms bidding on AdWords does exactly this, and classical economics predicts he is right.  With the information I gave Google to sign up for this option, there doesn’t even need to be a competing firm for them to intuit how high my maximum price is.

But, for the moment, I’m extraordinarily happy, as my advertising has gone from barely above water to, what is the vernacular, making money hats.  (Lets see, 25 conversions a day is 750 a month times 2.5% is about 20 additional sales for $300 profit.  Not bad for a business which typically profits about $700 a month.  Not bad at all.)  I’m really praying the preliminary results hold up.