It isn't obvious

Consider the following keywords
You would like to bid each keyword so that you maximize orders subject to the constraint that the overall Cost Per Order (CPO) is less than $20. A lot of us would simply look at each keyword and bid them to a CPO at or higher than the average CPO. So we bid keyword 1 to position 1 (CPO of $20) and keyword 2 to position 2 (CPO of $20). In this scenario we would get 15 orders from keyword 1, 10 orders from keyword 2 for a total of 25 orders at a cost of $500 ($300+$200). The overall CPO is $500/25=$20. So it's good.

But what if we did the exact opposite? Here we would get 10 orders from keyword 1 at a CPO of $10 and 20 orders from keyword 2 at a CPO of $25 for a total of 30 orders at a cost of $600 ($100+$500). What is the CPO now? Its $600/30=$20. We got more orders for the same CPO despite exceeding the CPO target for keyword 2! Welcome to the portfolio theory of SEM. Like we said... It isn't obvious.

Learn more about the portfolio theory of SEM and several other cutting edge SEM concepts in our latest paper: Algorithms and Optimization

Comments

[Kevin]

Retailers generally set CPO goals not arbitrarily. When they say they want their CPO goal to be $20, that doesn't mean some at $25 are ok as long as the average is $20. Many retailers would be losing money here.

[Efficient Frontier]

That is a very valid point Kevin. However, there are two points to consider:

1. Many retailers want their search campaigns to spend to a budget with a CPO constraint. This is a volume play. By constraining every product/keyword to a CPO you might control your profit but you might not be able to spend to your budget. The portfolio method ensures that you spend to your budget at or lower than your overall CPO goal in the best possible way.

2. A second situation arises when an advertiser may want to bid every keyword to the position where profit is maximized (profit maximization). I believe, this is the situation you are referring. A model based portfolio method, such as Efficient Frontier’s bid optimization method, will predict the profit at every budget level to a high degree of accuracy. The point where the profit at a keyword and therefore a portfolio level is maximized is the budget to be set. In this situation, the portfolio based method has two additional advantages:

a. Although each keyword is bid to individual CPO goals (implicitly), it tells you the overall spend and ROI beforehand. This lets the advertiser plan her budget and also gives her a high degree of visibility.

b. It also gives you the trade offs you are making by bidding each keyword to the point of maximum profit. For example: For a 10% increase in profit would you be fine with a 8% decrease in volume?

I will end with another example:

You have been given a $100k budget to spend in a month. Lets assume you have accurate keyword models. You find out that by bidding every product to its CPO you can only spend $80k. In this situation, you would do one of two things:

1. Spend only 80k that month.

2. Try to figure out where to put the additional $20k so that you get the highest possible revenue. This can be very hard if you have 100k or so keywords as there are many keywords and positions to be considered. For a 100 keywords at 8 positions, there are billions of possibilities (8^100).

The model backed portfolio method can not only do both with more efficiency than a human or a simple heuristic, but also give you visibility into the trade offs made at every spend level.