One other difference between ads and Wall Street: each impression is worth a different amount to different advertisers. An impressions from a teenage girl is worth more to a fashion brand than an enterprise software company.

Equity traders might disagree on the proper valuation for a stock but they are all anticipating the same payout: future cashflows. There might be differing opinions on how to discount the future cash flows, just as there can be differing opinions on the value of the teen girl impression to the fashion company, but the true value of a stock doesn’t depend on the buyer’s identity. Different from stocks, the value of each impression does vary from brand to brand.

Mike — looking forward to the cookie ID syncing post

I just wanted to say thanks for hosting such an insightful blog that goes so deep into the technical workings of the advertising ecosystem. It’s rare that you see anyone giving as much effort as you do to help others understand.

As for RTB, it sounds like a very promising technology that we all look forward to.

- Peter

Great comment.

- Not sure I agree about a battle between server based & browser-based. Yes, there’s a huge hurdle to doing server based RTB integrations today (server-side cookies), but I don’t think that’s going to last for long. Also, there are a variety of mechanisms to sync cookie IDs that address the proprietary user information piece. (worthy a blog post I guess!).

- I’m anxiously awaiting the day where better data and faster trading make a difference! Agree RTB is the beginning of this world actually happening.

I think there are going to be a lot of companies vying to be the Goldman & Sachs of online advertising. I think AOL is terrifically positioned with the technology and people from Ad.com to be in this position. Of course they could also turn into Lehman Brothers =).

-Mike

– There’s going to be a very interesting battle between server-based RTB and browser-based RTB. Obviously, Google and RM much prefer the server based model (more control, more bidders), however buyers who have their own proprietary user information in cookies will be able to utilize that info at decision time in a browser based model.

– As wall st. taught us in the last 20 years, the way to sustainably profit doing algorithmic trading is based on having better data (to price accurately) and trade faster. It will be of particular interest to see the data landscape develop, between open data exchanges (ala bluekai, exelate) and proprietary data based ‘hedge funds’.

– The analogy to wall st., however, breaks in several important ways that people excited about RTB don’t realize many times. First, unlike (most of) the objects sold in wall st., media is perishable. That changes fundamentally the value of demand. Second, what comes inside bought ad placements are ads. So, even if one can segment ad placements up to the 10th digit, if at the end there are only two creatives to choose from, the potential benefit is highly depressed. That’s why the advancement of real time bidding will also bring a corresponding exponential growth in the importance of dynamic ads, which I dare to predict will become mainstream, at least in the DR market in the next couple of years.

One final anecdote, I asked Tim Armstrong this week how he thinks the advancement of exchanges and RTB will impact the ad network business of AOL. His answer was, if exchanges are nasdaq, we at AOL want to be Goldman Sachs…

Keep it coming,
Eran