Marginal ROAS: How to Know When to Stop Scaling on Cyber Monday
Marginal ROAS is the return on your next dollar of ad spend, and on Cyber Monday it is the number that decides whether scaling makes you money or just makes you busy. Average ROAS for the day can look excellent while the last $10,000 you added returned well under break-even. Cyber Monday falls on November 30 this year, and Adobe expects it to be the first U.S. day to pass $15 billion in online spend. That much demand makes it tempting to keep raising budgets. This post is about knowing when to stop.
Average ROAS vs marginal ROAS
Average ROAS is total revenue divided by total spend. Marginal ROAS is the extra revenue from an extra increment of spend, divided by that increment.
Google's Meridian documentation puts the distinction cleanly: ROI is "a historical, channel-wide average," while marginal ROI is the incremental outcome on your next dollar above the current budget level. One grades the past. The other tells you what happens if you push.
A simple Cyber Monday example on one Meta campaign:
- $20,000 spend, $80,000 revenue: average ROAS 4.0
- Raise to $30,000, revenue $96,000: average ROAS 3.2, marginal ROAS on the extra $10,000 is 1.6
- Raise to $40,000, revenue $104,000: average ROAS 2.6, marginal ROAS on the extra $10,000 is 0.8
At $40,000, the day still reports a 2.6 ROAS. It looks fine. But the last $10,000 lost money on revenue alone, before discounts, COGS, and shipping. Average ROAS hides that because it blends the cheap early conversions with the expensive late ones.
Why diminishing returns on ad spend hit sooner over BFCM
Diminishing returns arrive faster during BFCM because the cost of each impression rises while the pool of high-intent shoppers you can reach does not grow as fast.
Billy Grace's 2025 data showed median CPMs up about 36% in Black Friday week, with Meta and TikTok up roughly 45% and Snapchat up 72%. Conversion rates rose too, which is why cost per order still fell on average. But averages are exactly the problem here. The first tranche of Cyber Monday spend reaches people who were already planning to buy. Each step up in budget pushes delivery into colder, more expensive audiences.
Recast describes the shape most channels follow as a concave curve, strong early impact that flattens as spend rises, and notes plainly that "you cannot double spend and expect sales to double." On peak days the curve flattens earlier, because the auction is crowded with every other brand doing the same thing.
There is also a measurement wrinkle. During Cyber Monday, returning customers buying from email and organic search can get claimed by paid channels through view-through and engage-through windows. Platform-reported ROAS stays high while the true marginal return falls. That is why the spend-step check below should run on click-based, first-party revenue, not platform numbers. We explain the gap in why Facebook over-reports.
AdBeacon reports click-only, first-party revenue by campaign and ad set next to platform-reported revenue, so Cyber Monday budget steps rest on numbers you can verify.
How to spot diminishing returns with spend-step data
You do not need a media mix model to estimate marginal ROAS on Cyber Monday. You need budget changes in deliberate steps and revenue read at each step.
1. Set your break-even marginal ROAS before the day
Calculate it from your Cyber Monday margin. If your discounted contribution margin is 35%, break-even ROAS is about 2.9. Anything below that on the marginal dollar loses money on the first order. If first orders are acquiring new customers with strong repeat behavior, you may accept a lower floor, but decide it now.
2. Scale in steps, not in a single jump
Raise budgets in increments of 15% to 25% at set times, for example 6 a.m., 10 a.m., 2 p.m., and 6 p.m. Eastern. Large single jumps make it impossible to tell which dollars did what.
3. Log spend and click-attributed revenue at each step
A simple table per campaign works:
- Step, cumulative spend, cumulative click-attributed revenue
- Incremental spend since last step
- Incremental revenue since last step
- Marginal ROAS (incremental revenue divided by incremental spend)
4. Compare against the same hours on Black Friday
Time-of-day effects are real. Revenue naturally rises into the evening. Comparing the 2 p.m. to 6 p.m. window on Cyber Monday against the same window on Black Friday, at similar spend, gives you a rough baseline for what was organic momentum versus what the extra spend bought.
5. Stop when marginal ROAS crosses your floor twice
One weak step can be noise. Two consecutive steps under break-even is a signal. Hold budget there or step back, and move the next dollars to a channel whose marginal ROAS is still above the floor.
A Cyber Monday ad strategy built on the next dollar
A good Cyber Monday ad strategy treats each channel as its own curve. Google Shopping and branded search often have lower CPM inflation and stay efficient longer. Meta and TikTok may saturate earlier. Shift money toward whichever channel still shows marginal ROAS above your floor, instead of scaling the one with the best average.
A few practical guardrails:
- Apply a Recent Buyers Exclusion audience to prospecting so scaled spend is not chasing people who just ordered.
- Watch the new customer share of orders as spend steps up. If it falls while spend rises, the extra budget is mostly reaching returning buyers.
- Do not chase late-night spend just because average ROAS looks strong. That is often where marginal returns are weakest.
For the longer view on scale decisions, see when to scale your Facebook campaigns.
Where AdBeacon fits
AdBeacon reports click-only, first-party revenue by campaign and ad set next to platform-reported revenue, so you can run spend-step checks on numbers that are more verifiable than self-reporting. For the longer-range view, AdBeacon's Google Meridian MMM integration, which stays free, models response curves and marginal ROI by channel from your historical data. Luma, AdBeacon's AI assistant built on Claude, can flag when a campaign's returns drop after a budget change.
Closing
On Cyber Monday the average will almost always look good. The useful question is what the next dollar earns, and you can answer it with planned spend steps, a margin-based floor, and click-attributed revenue. If you want to see how your channels responded to budget changes last BFCM before you set this year's plan, book a live AdBeacon demo.
FAQ
What is marginal ROAS?
Marginal ROAS is the additional revenue generated by an additional amount of ad spend, divided by that additional spend. It measures the return on the next dollar rather than the average return on all dollars.
How is marginal ROAS different from average ROAS?
Average ROAS blends every dollar spent, including the cheap early conversions. Marginal ROAS isolates the most recent spend increase, so it falls much faster than average ROAS as a campaign saturates.
How do I calculate marginal ROAS during Cyber Monday?
Increase budget in planned steps, then divide the revenue gained since the last step by the spend added since the last step. Use click-attributed first-party revenue so platform view-through credit does not inflate the result.
When should I stop scaling ad spend on Cyber Monday?
Stop scaling a channel when its marginal ROAS falls below your break-even ROAS for two consecutive spend steps. Move the next budget to a channel whose marginal return is still above that floor.
Why do diminishing returns on ad spend appear sooner during BFCM?
CPMs rise sharply during peak week, and early budget captures the highest-intent shoppers first. Each additional dollar reaches colder, more expensive audiences, so returns flatten earlier than usual.