For those in the industry, this graph has been making the rounds today.
You can find Pubmatic’s own view of it here:http://www.pubmatic.com/adpriceindex/index.html

Everyone has their opinions of this; some say this is bad sign for Pubmatic, others say its a sign of the economic times.

So what is my opinion? I think the data just isn’t clear enough. You cannot convince me that the cratering of cpms from .38 to .18 in single month is actually a real drop. Do you really think that the large sites with the dedicated work force could not sell half of what they had been able to sell only the previous month? This just says to me there is something wrong with the data. My suspicion would be some new large users who are low performance sites. A single large publisher with low cpms using the system can easily cause this drop without adverse affects on others in the column.

The interesting piece is the 1.29 number. How is this coming about? Why is it so much higher than large sites? I don’t buy the idea that its because advertisers are buying niche. Quite frankly, 1) advertisers are still quite weary of small sites and more importantly 2) there is just not the capability for an advertiser to target niche sites they are looking for.

So, before people read into this graph too much. Remember, while 3,000 seems like a large number, it is a tiny fraction of the internet. To make anything other than broad judgments on this data is overreacting.

Just read the most recent Seth Godin post

Essentially, he says that Firefox users are a different breed than IE users, and as such should be treated differently. These people are the ones who went out of their way to download a different browser. He likens them to people who went to college.

I can’t help but first think what this means for internet advertising.

The Firefox users, while being a key demographic for many advertisers, are also the ones least likely to click on ads. They are the ones most likely to have extensions which block ads. Essentially, these are the users who in the current context of internet advertising are worth the least to a publisher. If you are going to “treat them differently” as Seth suggests, should it be treating them better or worse?

Now the hard part. How do you fix this? It is a big problem, as Advertisers right now are very much into hard metrics, its one of the big upsides to advertising on the internet in general. In fact, if I was an advertiser right now, I’d try and target these people at low cost, knowing I’m throwing out brand awareness while paying out very little for these users (who are some of my most coveted demographic). The long term solution, ironically, may come from the out-of-favor CPM. Sales needs to convince advertisers that if they are looking at branding, low CTR does not make a campaign fail. The metrics do not explain the true value.