After Facebook ads part 1, I ran a few test ads that performed abysmally (although I’m targeting the kind of power users who block ads). The actual price paid seem to be vaguely proportional to the suggested price, but offering a tiny fraction seems to work just fine (I’m generally bidding less than a 1/10th of the suggested price and managing to exhaust my budget). Today I’m comparing suggested ad rates by state and age and again I’m assuming that the suggested prices are proportional to the average prices.
First off, I ran a query across all 50 US states, and the first thing that I noticed is that the implied click through rate is always .04% (you’re always encouraged to spend 2.5 x the CPM rate for a click — the raw data is available at the end of the post in Excel format).
50 data points means we can start talking trends, but none seem to jump out in the state by state data, there’s a slight discount as a state’s Facebook usage goes up (as I suspect a lot of the inventory is being bought nationally, so more ad spots for the same # of ads would drive down prices).
Strangely GDP per capita either doesn’t affect prices or affects them negatively.
A lot of that has to do with demographics, older users are less likely to use Facebook. Even when you account for their higher click costs, they are not a significant portion of the revenue:
But here’s the interesting takeaway, the implied click through rates for all ages are still 0.04%. Whatever direction you think it goes, if you don’t believe 14, 34 and 64 year olds all click on the same percentage of ads, there’s an arbitrage opportunity here.
Here’s the raw data: