A dear friend of mine Craig Danuloff told me in 2009 that he hated ROAS as a metric.

A decade of experience since has only validated Craig's opinion for me.

I profoundly dislike ROAS, and if you like ROAS chances are that I like you less. :)

It distresses me to no end that companies and agencies, all around us use ROAS as currency - inflating its value far, far beyond its minuscule value.

Wait. What's ROAS?

      Return on Ad Spend.

It is a naval-gazing advertising-centric metric.

It is not a business metric. It is difficult to apply it to all our marketing activities (ex: owned and earned). Surprisingly, it is not that easy to optimize for.

And, there are other reasons. But, since the nerve damage in my hand is proving to be really painful today let me type less and illustrate using a simple example.

You are incredible and have used a sophisticated media planning solution to invest across marketing channels.

Bravo.

The activity resulting from your Marketing investment will yield company revenue. The table with that data looks like this...

Even in this small table, that is a lot of data.

It is entirely natural that our leaders want short code to assess performance quickly.

Enter Return on Ad Spend = Revenue / Marketing Cost.

In short order, the enhanced table will look something like this...

OMG! So Cool! So Easy!

I have to admit, it does look easier to understand.

2.00 is easy to compare to 1.77 or 0.34.

My core worry is how misleading this can be, because of what it is not being considered.

ROAS is hiding a lot more than it is reveling. Ex: What contributed to the revenue... Quantity, cost, discounts and promotions, and more.

Another element is subtle, but important. ROAS also sub-optimally equalizes attention on big and small spends.

I've observed that people ignore the huge differences in the yellow column and the conversation is hijacked by Altavista's 6.83.

(I call this descaling, a pernicious problem.)

In order to get our leaders to make smarter decisions, where the opportunity to influence exists, I like a different metric to show marketing effectiveness:

      Cost Per Sale.

How much did it cost to sell one unit of product?

The table now looks like this...

Surprising, right?

First, this is a higher bar for Marketing to hit.

Second, notice the patterns exposed now.

In some instances you see what you might expect. Ex: Washington Post and Altavista.

But, there are others you did not expect from the ROAS column. Ex: MySpace and Terra differences, or Yahoo!

Cost Per Sale brings a sharper lens on what matters to the company more, it frames effectiveness differently. It is not hard to imagine how much smarter the discussion will be off the purple column than the blue one.

Next time you are asked to compute ROAS, experiment with using Cost Per Sale... Sprinkle in some conditional formatting... An significantly more informed discussion will follow...