ROAS* is the Rodney Dangerfield of media metrics. Gets not respect…after all, who needs to know whether advertising has generated any business value?

And who needs a disciplined way to set the advertising budget when we already have the industry’s trusted planning methodology: “Hey, let’s plus that thing up a bit this year.” Or, “Sales look like they are going to be soft, so cut back a bit”.

Besides, everyone knows the best way to kill good advertising is to hold it accountable.

So yes, ROAS deserves the contempt it receives.

And pay no attention to those math and economist nerds who point out that ROAS is the calculus of advertising effectiveness. It is literally the rate of change in sales as advertising spending changes:

ROAS_marginal = dS/dA

The budgeting rule most are not using.

Suppose your gross margin rate is 50%. Each additional dollar of revenue contributes 50 cents toward profit. Therefore, the final advertising dollar should generate $2 in incremental revenue. That produces $1 in gross profit, exactly covering the $1 cost of advertising. Take longer term sales result and a lift in brand favorability as gravy.

In other words, the profit-maximizing spending point occurs when:

ROAS_marginal = 1/(Gross Margin)

…The economic stopping rule for advertising investment.

At a 40% margin, the required marginal ROAS is $2.50. At a 25% margin, it is $4.00.

Although hey, I must admit that DOES look incredibly useful!

The supposed small-spend hack

ROAS critics will tell you that the metric isn’t valuable because it can be gamed by spending less. And they are correct. If you cut spending until you are buying only the easiest, highest-return opportunities, measured ROAS will generally rise.

Ah, but that is a self-inflicted wound.

Study the equation below, where advertising elasticity (the thing on the left) is measured by almost everyone via marketing mix modelling:

Advertising elasticity= ROAS_marginal × A/S

A is advertising spending and S is sales.

Rearranging it gives us the famous Dorfman-Steiner budget setting equation:

A/S = advertising elasticity divided by ROAS_marginal

ROAS and advertising elasticity determine the economically supportable ratio of advertising spending to sales.

Yes, you can make ROAS rise by driving down the advertising budget. But which CMO wants to make the case that they need a smaller ad budget?

Instead, set the minimum marginal ROAS required by the economics of the business. Then continue spending along the diminishing returns curve until the last dollar reaches that requirement.

Get every ounce of juice from the squeeze.

Are you underspending and short-changing growth? I bet you are.

I have a strong suspicion that few marketers follow these simple equations. Instead, they follow precedent. And my guess is that marketers are more likely than not to be underspent and operating below the growth curve their brands have the potential to deliver.

Once incrementality is measured correctly, ROAS answers the question executives have every right to ask: What profit boost did we receive for the money we spent?

And Now, Enter Movable Middles©

This is where the power of the MMA’s Movable Middles becomes even more decisive.

Target Movable Middle consumers who are more responsive to advertising and incremental ROAS rises. Once that happens, the advertiser is above its required return threshold again.

That creates room to increase the total budget while continuing to meet the company’s ROAS requirement.

Better targeting should warrant a bigger budget, leading to more growth. That is a virtuous circle rather than a vicious one.

I am testing a skill I built using ChatGPT that follows the Dorfman-Steiner optimality equation to optimize total advertising spending and extended to audience allocation. Across a series of what-if exercises for a number of brands, optimizing the allocation of a fixed budget seems to have the potential to increase incremental sales by roughly 10% to 50%.

And that is just the first-order opportunity.

When the model also increased the total budget until marginal ROAS again reached the required hurdle, the absolute incremental-sales gain often doubled or tripled again.

ROAS allows an advertiser to estimate diminishing returns, compare the productivity of audiences, reallocate spending and determine whether the total budget should rise or fall.

All terrible stuff…no respect at all.

*ROAS is a calculation of return on advertising that is defined as the incremental revenue generated by advertising, divided by that increase in ad spending. Some call this iROAS differentiating it from ROAS where all sales are divided by all ad spending but to me, that is a fairly useless calculation.