How to Handle Declining ROAS

How to Handle Declining ROAS

Jan 25, 2021 David Deppner

Let me guess… Your ROAS was strong, massive even, but now it’s sinking and making you wonder what in the heck went wrong. Does that sound familiar? This is one of most common concerns we get from ecommerce companies. Results have been sliding for many merchants who are now wondering if it’s even possible to get a higher ROI in the current environment. But there’s a different perspective to have here that they’re missing. Let’s dive in.


First off, when people raise these issues, it's usually because they were achieving very high Return on Ad Spend (ROAS) ratios at some time in the past, and those ratios have dropped more recently. Digital marketers commonly refer to this ratio of returns to ad spend as "ROI" although that is technically inaccurate. For the purposes of this article, I'll use "ROAS" and "ROI" interchangeably since they are the common industry terms, although I think it will become obvious that these metrics are distractions from the real underlying economics of how we achieve profitability in digital advertising.

Your ROAS Doesn't Matter

There are several quick answers I could give about what to do with declining ROI, and one is that the ROI you’re calculating simply doesn’t matter. Let's explore a quick example.

Let's say you have 50% margins. If you spend $1 to get $20 in sales, you end up with $10 in gross profit. And after subtracting the $1 in ad spend, you have another $9 in profit that you didn’t have before. You should buy a lot more of those clicks. They’re highly profitable.

A key characteristic of ad auctions is that, as you try to ramp up your spend, the returns will drop. So if you spend $1 and get $20 in sales, spending another $1 won't get you another $20. It's more likely to get you less. As you spend more, and your marginal returns drop, what really impacts your total profit is what is happening with each additional dollar you spend.

If your returns decline to the point you’re spending $1 to make $3 in sales you’re still making a lot of extra money at 50% margins. The $3 in sales is bringing in an additional $1.50 in gross profit, which puts an extra 50 cents in your pocket after accounting for the $1 in ad costs.

I want to be clear here. In this example, you’re making a profit on the 20:1 ROAS portion of the results and on the 3:1 ROAS portion… and everything in between.

If you took this deal and looked at the measured ROAS in the ad platform, what you'd really see is that the average ROAS is somewhere between those two figures. The average ROAS/ROI doesn’t matter. What matters is whether you’re still profitable on the portion of your results that are getting lower returns, given what you're spending on ads, and what your returns look like. Does spending $1 more bring you additional profit or does it not? The average is not useful for making any financial decisions at all. Only the marginal results matter because you’re getting a lot of clicks for cheaper than that which are even more profitable than the results at the edge.

In this example, it is profitable to take the 3:1 ROAS slice of sales, and even to take some additional sales beyond that, approaching 2:1 ROAS, because at 50% margins, you will get additional profitable sales all the way until your marginal returns hit a 2:1 ROAS. If you can spend $1 at 3:1 ROAS and net out an additional 50 cents on top of getting your dollar back, what rational person would turn that down?

In the real world, there are very few investments that pay an instant return of 50%.

Profit is What Matters

If you want to optimize for total profitability, then you have to stop thinking of this as an average income vs. expense ratio and instead approach it from a finance and microeconomics profit maximization perspective. The real underlying reason the ratio doesn’t matter is that if you’re profitable at the margins then, every time you make a sale, you recover the money you've spent since the previous sale. Don’t think of it as spending some large amount every month. Think about it as investing a small amount, getting it back with profit fairly quickly, and then doing it again.

Let's explore another example where I get 240 sales in a day and my average ad cost per sale is $10 with an average sale of $30 at a 50% gross margin. Am I really spending $2400 per day and $72000 per month? Sure, from an accounting perspective. But that's not the only way to view it.

It’s just as valid to say that I’m really only ever putting about $10 in ad cost at risk before I make the next sale. Every 6 minutes, I spend $10 on ads, and every 6 minutes I get about $30 in sales, with $15 of that in gross margins. Of that $15 in gross margins, $10 returns the money I just spend on ad cost, and I'm left with a net of $5 I didn't have before.

In this example, about every 6 minutes, I part with $10, get it back again, and along with it comes an extra $5 in profit I didn't have before.

From an accounting perspective, that's a $72,000 expense every month. But there's only ever $10 at risk, getting recycled every 6 minutes, with a 50% return every time. Sign me up for some more of that. Perhaps that's not some massive expense, but just a tiny amount of working capital tied up in getting my next sale, that keeps getting reinvested after every sale.

Remember, the accounting perspective on money is rooted in a deep philosophical tradition of pessimism and conservatism. Do we care about how an accountant will code the transactions when filling out financial statements for tax accounting purposes? Or do we care about maximizing shareholder value by achieving the highest profit possible using a small amount of working capital?

High ROAS Means Wasted Potential

If you say that you used to get a really high 20:1 “ROI” my immediate reaction is: Why didn’t you spend more and tolerate lower return ratios, but achieve more total profit? Optimize for what the business owners want more of—cash in their bank accounts. We have a client right now that is running between 20 or 30 to 1. They know they could drop that average to 10 to 1 and the marginal returns down to 2 or 3 to 1 and make far more total profit. But they have supply issues getting new stock into their warehouses, so we take higher return ratios on lower sales volume rather than trying to ramp up total profit. But that is a limitation on the product supply, and if they can get their stocks replenished, we’ll immediately start tuning them towards more total profit, which will have the side effect of dropping their average ROAS metrics dramatically as their profit rises.

Read that again. If you have a ROAS that is abnormally high, like this client, then you are leaving money on the table and you could earn far more total profit by dramatically reducing your ROAS. Many businesses are in this situation with reduced profits because of irrational ROAS goals that are set too high without any understanding of how that hurts the business.

What Does Declining ROAS Mean?

What about the broader general issue of declining average returns that so many people see in their digital advertising? One answer is that other firms have adopted the same bidding strategies that you have. More merchants are making the same decisions, based on the same or similar data. Of course that increased competition is going to drive up costs on the profitable clicks. That’s the beauty of an auction, if you’re Google. Competing advertisers will tend to drive the price per click up to the maximum that the market can bear, and the seller (Google) wins.

With anything that’s highly profitable, competitors will bid the prices up, until you’re right on that edge squeezing out the profits. The only solution is to become more sophisticated than your competitors and squeeze out any little bit of performance that gives you an advantage. If you only do what everyone else does, you’ll settle for an average level of results that everyone else is getting as well.

Consider this carefully when simply adopting "best practices" and not actually customizing those practices to the nuances of your own firm and your particular competitive advantages. If you just do what everyone else is doing, you will only get average performance, with declining profit over time as more competitors pile on and do exactly the same thing.

Find Your Competitive Advantage

If you are going to use the same ad platforms and same campaign types and same automated bidding strategies that your competitors use, then you will need to work like crazy to make other inputs to these systems better than your competitors. For example, you need to make your shopping feed data amazing. Do you have perfect data in every way? Did you already deploy the latest feed attributes that Google just added (product_detail and product_highlight)? Do you have full GTIN data on all items? The only way to outperform other people is to understand the competitive landscape better and to more effectively take advantage of what opportunities you can, given the reality of the things which you can't improve.

If you have a complex mix of products with varying gross margins and order values, then you need to take a more nuanced approach rather than throwing all of your products in one campaign and setting an average ROAS target and evaluating success against that target. You're better off not using ad platform automated bidding at all in some circumstances, because they're generally trying to optimize for maximum revenue or a given ROAS and different products with different gross margins contribute different amounts of profit that these bidding strategies are unaware of. Many ecommerce sites can outperform AI bidding algorithms by using good old-fashioned Max CPC bidding if they have varying margins, because of this issue, with appropriate attention paid to the underlying financial models. Automated ROAS-based bidding algorithms will tend to overbid on the items you sell with the worst margins and drag down the proportion of your returns that is actually profit remaining after ad cost.

As everyone else moves in lockstep towards ROAS-based automated bidding, those firms that take a more nuanced approach can push more total profit. Everyone else is making the same predictable mistakes. By understanding those mistakes, we can buck the trend and pull in money that they are leaving on the table.

Try More Things; Try Newer Things

As competition increases on the ad platforms and strategies that you've used in the past no longer work, look to more recent opportunities in order to find things your competitors aren’t taking advantage of. It's amazing how many advertisers are doing precisely what they were doing 5 years ago, with minimal changes. That was a lifetime ago in this industry.

But above all, stop making decisions based on average ROAS! You can tolerate lower ROAS on your most profitable items, and you shouldn't tolerate average ROAS on your least profitable items. Total profit after all expenses including ad costs is what matters most, not average ratios between income and expense. Adopt a more sophisticated perspective on achieving a portfolio of results with different levels of returns across your product lines, based on their contribution to actual bottom-line profit. Profit matters, not ROAS.


Psyberware specializes in managing online advertising for ecommerce businesses, and nothing else. If you'd like to work with an agency that truly understands how to achieve profit in the current environment, get in touch with us.

Previous Post Next Post