Portfolio Theory is another term, along with ‘Arbitrage’, that I think maps very poorly to the search world.
Efficient Frontier, a bid-management and search marketing firm, has been using it for years in their marketing.
I think the analogy breaks down because portfolio theory holds the inherent concept of diversification. Diversification reduces risk and hopes to hold the same expected return.
A marketer bidding on a ‘portfolio’ of keywords isn’t diversifying. They are essentially picking one category of keywords with one goal in mind. There is no concept of asset diversification, no concept of diversification amongst industries.
But that’s not to say Efficient Frontier isn’t doing great things. They are. Clients love them. Just that it has nothing to do with portfolio theory.
The take away, in my mind at least, is they are saying to marketers ‘don’t give up on keywords too early’ and ‘don’t get overly excited about success in small samples’. Both great advice points.
Essentially: “Don’t let the data fool you”.
To make better decisions you need statistically significant sample sizes. They help clients by trying to get data scale on long tail terms by aggregating response and conversion results across past and current clients. That has real value.
Long tail terms by their very definition are infrequently searched and in the scheme of things, advertisers have even less data around click-through rates and even less data that that on conversions.
So to me, scale is the issue around decision-making. And they are helping clients make better decisions through their institutional knowledge. Long tail terms by their very definition will need longer time periods (perhaps years) to be correctly valued and priced. But if they are storing and learning, over time they will be able to help and guide newer search marketers who do not have that marketing memory.