You raise the budget and ROAS, how many dollars in sales each dollar of ad spend brings back, drops. The almost automatic reaction is to assume something broke. Most of the time, that's not what happened: it's the law of diminishing returns, a general economic principle that also applies, almost to the letter, to digital advertising. Understanding it replaces the wrong question ("why did my ROAS drop?") with the one that actually matters ("am I still making more money even though ROAS is lower?").
What the Law of Diminishing Returns Means, in Plain Terms
It's a simple principle: when you keep adding more of the same resource (in this case, ad budget) while everything else stays fixed (the size of your market, how many people could be interested in your product), each additional unit of that resource yields a little less than the one before it. It's like watering a plant: the first glass of water helps enormously, the fifth already adds less, and past a certain point, adding more water doesn't just stop helping, it can start drowning it.
The same thing happens with advertising audiences: the first dollars of budget reach the people most likely to buy (the most receptive, the ones most similar to your current customers). As you raise the budget, the platform has to go find increasingly less aligned people to spend that extra money on, which is why each new dollar returns less than the previous one.
Why This Is Expected, Not an Account Error
This isn't a flaw in the campaign or in whoever manages it, it's the mathematical reason behind phenomena we've covered in other articles: the partial algorithm learning reset and audience saturation explained in how to scale on Meta without losing ROAS, or the choice between going deeper or going wider covered in horizontal vs. vertical scaling. Both of those articles cover the "what to do"; this one covers the "why it happens."
The Average Hides What Matters: Marginal ROAS
When you look at the ROAS on your total monthly spend, you're looking at an average. But what actually determines whether scaling further is worth it is marginal ROAS: how much you generated with just the last chunk of budget you added, not with the total. An average that still looks good can be hiding a final tier that's no longer profitable, and conversely, an average that dropped can still be a better business than before, in real dollars.
A Numerical Example
Illustrative example, not based on a real business. A business with a 40% gross margin tests different levels of monthly budget:
| Budget | Sales Generated | Average ROAS | Profit (sales × margin − budget) |
|---|---|---|---|
| $100,000 | $500,000 | 5.0x | $100,000 |
| $200,000 | $900,000 | 4.5x | $160,000 |
| $300,000 | $1,200,000 | 4.0x | $180,000 |
| $400,000 | $1,400,000 | 3.5x | $160,000 |
Notice what happens: the average ROAS drops at every step (5.0x → 4.5x → 4.0x → 3.5x), but profit in dollars rises all the way up to a $300,000 budget, and only drops at the last step. That's the answer to "when does scaling with a lower ROAS still pay off": as long as total profit keeps rising, a lower average ROAS isn't a problem, it's the logical cost of selling more. The problem shows up when, like in the last step, profit starts falling even though ROAS still looks "acceptable" at a glance.
If you calculate the marginal ROAS of each added tier (how much that specific last budget increase generated), the limit becomes much clearer: the final tier (from $300,000 to $400,000) generated only $200,000 in new sales from $100,000 in new budget, a marginal ROAS of 2.0x, below this business's breakeven ROAS (1 ÷ 0.40 = 2.5x). That specific tier was no longer profitable, even though the overall average still read 3.5x.
How to Find Your Optimal Scaling Point
- Calculate your breakeven ROAS based on your real margin, it's the floor, not the goal.
- Look at the marginal ROAS on every budget increase, not just the cumulative average: new sales generated ÷ new budget added.
- Keep scaling as long as the marginal ROAS is above breakeven. As soon as the marginal figure drops below it, that specific tier of budget is subtracting value, even if the overall average still looks healthy.
- Scale gradually, as recommended in horizontal vs. vertical scaling and how to scale on Meta without losing ROAS , that's what lets you measure the marginal return of each step instead of blending everything into an average that no longer distinguishes anything.
Common Mistakes
Looking only at average ROAS. A healthy average can be hiding a final budget tier that's no longer profitable.
Stopping scaling the moment ROAS drops, without checking profit in dollars. If total profit keeps rising, a lower ROAS isn't a problem, it's the expected cost of selling more.
Continuing to scale "because the average still looks fine." The average reacts slowly; the marginal figure warns you sooner.
Frequently Asked Questions
Is it bad for ROAS to drop when I raise my budget?
Not necessarily. It's mathematically expected due to the law of diminishing returns. What matters is whether total profit in dollars keeps rising, not whether the ROAS ratio stays the same.
What is marginal ROAS?
It's the return generated specifically by the last tier of budget added (new sales divided by new budget), as opposed to average ROAS, which blends all spend for the period together.
How do I know when to stop scaling?
When the marginal ROAS of a new budget increase drops below your breakeven ROAS, that specific tier is no longer profitable, even if the overall average still looks fine.
Does the law of diminishing returns apply to any business?
Yes, it's a general principle of any market with a finite audience, how quickly it shows up varies with market size and brand differentiation (see brand vs. performance marketing), but the underlying pattern is the same.
Conclusion
A ROAS that drops as you scale isn't, by itself, bad news, it's what's expected when you add budget to a finite market. The question that actually determines whether it's worth continuing to scale isn't "did ROAS drop?", it's "is the last tier of budget still leaving profit above my breakeven point?"
Not sure if your lower ROAS is still profitable? Book a free audit with KLIV and we'll calculate your real marginal ROAS together. Book a call →