So I'll just say that I agree with him in some major areas, but think he is missing some important things too. Definitely worth reading if you are interested in these things. Also if you are curious about what I work on, since the "CTR * price instead of CTR auction" is the kind of thing I work on now - both understanding the auction, and thinking about how to tweak it to improve efficiency.


Comments
There's a whole blogger policy at Y that covers what we are allowed to say, and I'm definitely within that.
On a somewhat related topic, it's interested to see the sorts of papers that show up in conferences from Y, G, and Microsoft. Google releases basically no papers on core search issues, Y does some but not many, and Microsoft has an amazing set of really great papers that give away everything they have. It's definitely not a coincidence that openness perfectly inversely correlates with winningness.
One would think that would be enough, but I don't think that is Google's policy. One problem is that my slant and analysis of even publicly available materials is colored by my inside knowledge. There are a lot of reasonable things one can say based on public knowledge, but knowing which ones an insider agrees with is still pretty meaningful.
And another problem is that G is just paranoid :).
On a somewhat related topic, it's interested to see the sorts of papers that show up in conferences from Y, G, and Microsoft. Google releases basically no papers on core search issues, Y does some but not many, and Microsoft has an amazing set of really great papers that give away everything they have. It's definitely not a coincidence that openness perfectly inversely correlates with winningness.
Yeah, I've noticed that :).
It is kind of sad, actually.
It is really awkward to be social with folks who are hugely enthusiastic and passionate about their work, who spend 40-80 hours a week on it, and then aren't able to talk a bit about what they love other than the food.
It is easy to see why so many Googlers seem to fall away from having an outside social life. Google provides one - with lots of great folks on the inside. It is a perk at least equivalent to the free meals.
I'm glad you are an exception, and I know there are many. But the Google social black hole is way too real and common around the bay area.
- chris
In the traditional model, advertisers find some target audience, and attempt to drown out other advertisers to get their message across. There's a tragedy-of-the-commons problem in that a more obnoxious ad is more likely to get attention, so these channels are full of largely uninteresting noise.
In contrast, with Google's search ads, there's a predetermined amount of advertising, and it is in Google's interest that the amount and presentation not be annoying. That is, instead of rent-seeking, advertisers are doing wealth-transfer (to Google) to get their message across, and since Google's interests are much more closely aligned with the searcher/consumer than in the older model, being advertised to is a much more pleasant experience.
I'd describe the key difference as being one of push advertising (traditional) vs. pull advertising, where the consumer more or less decides to see advertising, and gets what's most useful to them. Amazon's recommendations (while less tuned than they might be) represent a variant on the same principle: Amazon doesn't much care which of their products I buy, so their interest is in doing a very good job of recommending something I actually want (modulo profit margins on the items, etc.). In some sense, it's driving towards the utility of word-of-mouth, only without the unreliable transport/incompletely informed recommenders.
I should try to find my notes and blog the rest of it tonight.