How Does Google Ads Handle Rarity ?

How Should Google Ads Handle Rarity?

By Michael Knowles from MarketKing Consulting

Ok so you are selling a rare car. Let's say a 1963 Mustang. It's been preserved and conserved and lovingly cared for in the past 2-3 decades and today only gets driven around town on Sundays for show. Your asking price for the Mustang is $20,000. Your neighbour is selling a current day car, say a Ford Taurus, that is 8 years old. It's a family sedan that's racked up the usual 200,000 kms or miles. It's in OK condition, but you're tired of it and want to move on. The asking price for the Taurus is $12,000. You're setting up a Google Ad for this product - as is your neighbour for her Mustang.

Which ad should cost more on a pay per click (PPC) basis? The ad for the rare car or the ad for the common car? What would make sense to you?

Intuitively you might say that the owner of the classic car should pay more. How could we calculate a factor that attributes the cost of the PPC to the rarity of the product being offered? Say you live in a target area of 20 million car buyers. During that year 100 Mustangs of your model are sold and 100,000 units of the Taurus are sold in the same period.

We could consider a straight comparison - 100 Mustangs vs a population of 20 million represents 0.0005% of buyers, while Taurus sales for that model represent 0.5% of the buyers. How could we use this to adjust the Google Ads Pay Per Click to be charged?

We could improve the model by asking how many buyers really are in the market for the Mustang per annum? How many would genuinely consider it. Let's say that number is 10,000 buyers and the number of genuine buyers for the Taurus is 2 million. That increases the buyer potential ratio for the Mustang to a more realistic value of 1% and that of the Taurus to 5%. And that feels a bit more intuitive doesn't it, that the Mustang might be 5 times harder to sell than the Taurus - or maybe it is ten times harder?

So now we have a rarity factor (R) that could be used to calculate the relative PPC cost. Alternately we could consider the actual number of units of these cars that sell per year? If a total of 500 Mustangs are sold per year, that means 5 % of the market converts (conversion rate) and buys Mustangs, and there are currently 100 units for sale - the current market turns rate (or churn rate) is 5 times. The number of units for sale now represents one fifth of the annual sales. If 500,000 Taurus units of this model are sold per annum, then 25% of the market converts and there are currently 50,000 units for sale, then the market turns are a factor of 10 times per annum. The Taurus is twice as likely to sell at any time than the Mustang - and the Taurus should take 1.2 months to sell and the Mustang 2.4 months. 

Finally, let's assume that Click Through Rate on Google (the rate of conversion of the number of customers that see the ad to the number that click on the ad) is the same between the two sets of buyers. So if I was calculating a PPC for these products I might consider:

	Potential Buyers	Sell Price	On Market Now	Actual Sales Per Annum	Market Size $	Conversion Pct	Turns	"PPC"	CTR	Final Ratio
Mustang	10000	$20,000	100	500	10000000	5%	5	$5,000	0.01	$50
Taurus	2000000	$12,000	50000	500000	6000000000	25%	10	$30,000	0.01	$300

So perhaps in this case the Taurus seller should be paying six times more in PPC than the Mustang seller? Interesting to ask how this works for a house sale versus a book sale for example. Of course in this example, I have assumed that Google takes into consideration the actual sell price of the item, but in reality they don't. But maybe they should? What are your thoughts ?