Sunday, September 24, 2006

Findory switches from Google to Amazon ads

Findory recently launched a new version of our personalized advertising engine. This new version is based on Amazon Associates rather than Google AdSense.

Like the old engine, the new engine targets based on the content of the page and each reader's clickstream history on Findory. Unlike the old engine, it shows a targeted selection of books from Amazon rather than text ads from Google AdSense.

Why did Findory switch?

AdSense is an intelligent, self-optimizing, ad targeting system. It is a stubborn beast, convinced it knows what is right.

When Findory layered its own intelligent ad targeting system on top of AdSense, the two fought like crazed monkeys.

Findory would tell AdSense, "This page has articles about Google, Yahoo, search, engines, and technology," and then ask Google to target ads. Given that description, what do you think would be reasonable? Probably ads for web search engines and things related to web search engines?

Instead, Google sometimes would respond with ads for aircraft or automobile engines, blindly fixating on the word "engine" and apparently ignoring the rest of the information. It is hard to work with that.

We even tried test cases where we sent them nothing but a single keyword. For example, we said, "This page is about 'Google'." AdSense sometimes responded with ads for get-rich-quick schemes and penny stocks. That may be amusingly ironic. It even may be lucrative. But, it is not relevant.

After a year of experiments and optimizations to improve our targeting on top of AdSense, it became clear that we were not going to be able to bend it to our will. AdSense wants to target by itself. Any attempt to push it in one direction or another seems doomed to failure.

We decided to switch to a system where we would have more control. When we advertise books, Findory is completely responsible for the targeting. We analyze the Findory page and a reader's history, then our ad system picks specific books at Amazon based on that data.

Going back to that "Google" keyword test case, if I tell Findory's new advertising engine to target to the keyword "Google" (and nothing else), Findory responds with ads for four books: "The AdSense Code", "The Google Story", O'Reilly's "Google Advertising Tools", and "Google Maps Hacks". Ah, much better.

The new book ads target to any page on Findory in real-time as the content changes. Check out the targeting for Wired Magazine, ScienceDaily, Gizmodo, and Google Blogoscoped. And, of course, don't miss the targeting on your personalized Findory front page and how the ads change as you click on new articles.

The performance of the new ad system is roughly the same as the old Google AdSense system. Clickthroughs are much higher than what they were before but, because Amazon Associates has a much lower effective payment per click (about $.05), the incoming revenue is a little lower.

I like the new system much better. The ads are relevant and useful. We have complete control. And, I like helping people discover new books.

Update: Four months later, the results of this experiment are not looking as positive as I had hoped. Amazon Associates effective payment per click is closer to $.02 ($20 average price of a book * 6% associates share * 2% clickthrough-to-sale rate = $.024). AdSense generates $.10 or more per click, so our targeting of Amazon books would have to be x5 more effective or better just to make up for the difference in compensation, probably more like x10 to really be worthwhile. x5-10 improvement in targeting is a high bar, and we have not been able to get anywhere close to that yet.

Update: Eight months later, Findory switches back to Google ads. Unfortunately, the Amazon ads could not be made competitive given the effective payment per click.

11 comments:

Matt Cutts said...

Just fyi, when I did a search for Google, I got three Google books and one book called "The Pre-Foreclosure Property Investor's Kit: How to Make Money Buying Distressed Real Estate -- Before the Public Auction"

An Easter Egg, maybe? :)

Author said...

Greg, can't you do both Adsense and your own customized Amazon ads? I don't think Adsense license prevents you doing that (since you roll you own customization for Amazon).

Greg Linden said...

Hey, Matt! Thanks for stopping by!

Heh, that's pretty amusing. I was able to reproduce that on a news search for Google. The reason why is that the first several news articles are about Google's legal dispute in Belgium.

Just because the targeting is explainable doesn't mean that it's great. I agree that Findory's new targeting is not perfect.

What is true is that the ad targeting now is fully under our control. Before, when investigating why the targeting was off, we'd often hit the dead end of "because of Google." That's annoying because there is nothing we can do to improve. Now that the targeting is under our control, we have the ability to improve it.

Greg Linden said...

Hi, Sorin. That's a good idea!

For now, I'm not excited about putting that much advertising on Findory pages. Maybe at some point we'll stack both the Amazon and Google ads. Or maybe just show more Amazon ads. For now, I like the simplicity and sparseness of the four book short column of advertising.

Thanks for the suggestion!

Kevin said...

I'm very upset that Adsense hasn't really improved their backend relevance in over a year. They may have improve the internals but it isn't obvious from an external perspective.

What's your eCPM look like vs Adsense. Have you computed this yet? All that work will be for nothing if the eCPM isn't better.

I've thought of doing the exact same thing actually... cool that you beat me to the punch because this way I can see the result while being lazy ;)

Kevin

Greg Linden said...

By the way, Matt, a quick way to see Findory targeting ads only to the word "Google" is to do a web search for [Google] on Findory.

A news search for [Google] also targets to the content of the page (the articles in the search results) but, currently, a web search only targets to the search keywords.

Greg Linden said...

Hi, Kevin. The eCPM is slightly lower.

As I said, the clickthroughs are much higher, but Amazon pays much less per click. So, the revenue and the eCPM are both slightly lower.

Despite a good conversion rate (click -> buy rate), Amazon just doesn't pay that much for a sale (about 6%). If the conversion rate is 3-5% and the associates reward is 6%, total compensation on a clickthrough is 3-5% * 6%. The average price of a book is roughly $20. So, you can do the math, it is about $.05 per click. That is a lot lower than AdSense.

In the end, the eCPM is only slightly lower with the new ads because our clickthrough rate is so much higher. But it is slightly lower.

Nevertheless, it is a deal well worth making. The drop in revenue is minor and the targeting is now totally under our control. If we want to improve the relevance of the ads, it is now our job and our job alone to do so.

Aron said...

Adsense is great for publishers who want hands-free operation, but what are the choices out there for those that have the capacity for much more deliberate selection?

A quick browse of the Amazon affiliate program doesn't suggest a whole heck of a lot of participants, though I suppose the nature of books allows them to stand-in for a lot of possible topics. You wouldn't be able to beat a direct link to an exhorbitant fashion appliance in connection with a red carpet awards-show news story though.

Too bad there's not some business 2 portal advertising auction hub. Or maybe there is, and I don't know of it.

Nick Lothian said...

I wrote a thing like this for Amazon a while ago, too for my blog. Of course, I don't have your deep personalisation data, but it works okay for an amature.

See http://www.mackmo.com/nick/blog/tech/2005/7/24/Targeted_advertising.txt for details and http://www.mackmo.com/nick/blog/blogging/2004/11/20/MalcomGladwellsWriting.txt for a good example.

clam said...

I've also tried AdSense and Amazon affiliates and other affiliates for my web site at RoundedCornr.

The site is help web designers put rounded corners in their web pages. (Yeah I know it's stupid, but it does attract a few hundred visitors a day and good enough for me to experiment with different ad services.) Adsense clearly has trouble understanding the page (corner cutter?!, imvu?!), although it does learn enough from click-thrus that some CSS-related ads are now shown. At any rate, the eCPM is dismal. I try to cheer myself up by knowing that MySpace's CPM is even lower.

The ads from Amazon are always relevant, and this is just using code I get from Amazon. No customization there. Just like your experience though, I have made very little from Amazon. In fact, barely enough for a couple cups of coffee at Starbucks for the last six months.

The other affiliates (web hosting, photo library, etc.) didn't pay anything at all.

Surprisingly, asking for a "fee to support" the site generated as much revenue as Adsense. (Google Checkout is also pretty anal and won't let me ask for "donation" or "contribution") This despite the fact that the request is buried at the bottom of the page. Thanks for human kindness.

All in all, the total revenue pays for web hosting and a couple lunches a month. The income may be skewed by the fact that a big chunk of my audience is from outside US, but I think the relative effectiveness is pretty consistent with what I've heard from other people.

Joseph Bruno said...

"The ads from Amazon are always relevant"?

I was just looking at the Amazon.co.uk page for a mathematics book and was puzzled to find that the Sponsored Links section was asking me to Find Men at GayFriendFinder.com, Find Lots of Men at FriendFinder, or even visit SeniorFriendFinder Male Dating. Then I looked back at the title of the maths book I was thinking of buying: Leonhard Euler: A Man to Be Reckoned with.

"The ads from Amazon are always relevant"?