If you run paid social, you already track account-level ROAS. But that number tells you almost nothing about which video ad is actually paying for itself. Learning how to calculate ROAS on video ad creative — not just on your whole ad account — is what lets you kill losing hooks early and double down on the ones quietly carrying your revenue.
Account ROAS vs Creative-Level ROAS: Why the Difference Matters
Account-level ROAS is a lagging, blended average. It mixes your best-performing hook with three that are barely breaking even and one that is quietly losing money. If you only look at that top-line number, you can't tell whether this week's dip is a platform issue, a landing page problem, or one tired video dragging the average down.
Creative-level ROAS isolates the return generated by a single ad — one specific video, one hook, one edit — so you can compare it against your break-even point and make a keep-or-kill decision. This is the number that actually tells you what to produce more of.
How to Calculate ROAS on Video Ad Creative: The Formula
The formula itself is simple. The discipline is in how you pull the two inputs.
- ROAS = Revenue attributed to that ad ÷ Spend on that ad
- Example: a single video ad spends €420 in a week and is credited with €1,260 in purchases. 1,260 ÷ 420 = 3.0 ROAS.
- If your average order value and margin mean you need 2.2 ROAS to break even, that ad is profitable and worth scaling.
The complexity comes from where "revenue attributed to that ad" comes from, and over what time window. Get those two decisions wrong and the number is meaningless, even though the math is correct.
Step-by-Step: Isolating Creative-Level Revenue and Spend
- Go to the ad level, not the campaign level. In Meta Ads Manager, Breakdown by Ad or the Ads tab shows spend and purchase conversion value per individual creative. In TikTok and YouTube, use the ad-group or ad-level reporting view, not the campaign summary.
- Pick one attribution window and stick to it. Choose a single setting (for example, 7-day click, 1-day view) across every creative you compare. Mixing windows between ads makes the comparison invalid even if every other step is correct.
- Separate spend per ad from spend per ad set. If several creatives rotate inside the same ad set, the platform will still report spend and conversions per individual ad — use that broken-down number, not the ad set total.
- Set a minimum spend and time threshold before judging. A video that has spent €15 over four hours has no statistical weight yet. A reasonable starting rule to test: wait until an ad has spent at least your average order value multiplied by 3, or has run for a minimum of 48–72 hours, whichever comes later.
- Calculate ROAS for the same window every time. Pull revenue and spend for the same rolling period — last 7 days is a reasonable default — rather than lifetime numbers, since lifetime figures bury recent fatigue under early wins.
- Log the result against the creative, not the campaign. Keep a simple record (hook used, format, first-frame style, result) so you can spot patterns across dozens of ads over time, not just judge one in isolation.
Blended ROAS vs Creative-Level ROAS vs MER
Teams often confuse three related but different numbers. Here's how they differ and when each one is useful.
| Metric | What it measures | Best use case | Main weakness |
|---|---|---|---|
| Blended ROAS | Total ad revenue ÷ total ad spend across all platforms and campaigns | Reporting to the business, checking overall health | Hides which individual creative is working or dying |
| Creative-level ROAS | Revenue attributed to one specific ad ÷ spend on that ad | Deciding which hooks/edits to scale, pause, or iterate on | Can be skewed by attribution windows, low spend, or audience overlap |
| MER (marketing efficiency ratio) | Total revenue (all channels) ÷ total marketing spend | Sanity-checking platform-reported ROAS against real business revenue | Too high-level to guide creative decisions day to day |
A practical rule: use creative-level ROAS to decide what to test and scale, use blended ROAS to track account health week over week, and use MER as a reality check when platform attribution starts looking too good to be true.
Building a Simple Creative ROAS Tracking Sheet
You don't need a dashboard tool to do this well. A basic spreadsheet with one row per creative is enough for most small-to-mid accounts. Useful columns:
- Ad name / hook description (e.g. "problem-first hook, UGC style")
- Launch date
- Spend (7-day rolling)
- Revenue attributed (same window, same attribution setting)
- ROAS (auto-calculated: revenue ÷ spend)
- Break-even ROAS for your margin
- Verdict: scale / hold / kill / iterate
Review this sheet on a fixed cadence — weekly works for most accounts — rather than reacting to daily swings, which are noisy for anything spending under a few hundred euros a day. If you're also tracking how a creative's performance decays over time, it pairs well with a creative fatigue score, since a falling ROAS is often the first sign fatigue is setting in.
Common Mistakes That Skew Creative ROAS
- Judging too early. A video that's spent €20 isn't failing — it hasn't had a chance to succeed or fail yet. Set your minimum spend/time threshold before you look at the number, not after you don't like it.
- Comparing across different attribution windows. A 7-day-click ad will almost always show lower ROAS than a 7-day-click-plus-1-day-view ad. Comparing the two is comparing different things.
- Ignoring audience overlap. If two creatives run to overlapping audiences, the platform's attribution model may credit the same purchase differently each time it reports, making week-to-week swings look like creative performance when they're really attribution noise.
- Only looking at lifetime ROAS. A video that opened strong and has been declining for two weeks can still show a healthy lifetime average. Always check the recent rolling window, not just the all-time number.
- Forgetting margin. A 3.0 ROAS on a low-margin product may lose money; a 1.8 ROAS on a high-margin product may be very profitable. Calculate your actual break-even ROAS once, based on cost of goods, shipping, and fees, and use that as your bar — not a generic industry number.
What Counts as a "Good" ROAS for Video Creative
There's no universal good ROAS — it depends entirely on your margin, average order value, and how much of your traffic is cold versus warm. Instead of chasing a borrowed benchmark, calculate your own break-even point:
- Take your average order value.
- Subtract cost of goods, payment processing fees, and shipping cost.
- Divide average order value by what's left — that's roughly the ROAS you need just to break even.
- Anything above that, after accounting for returns and discounts, is contributing margin you can reinvest in scaling or testing.
Treat any ROAS target you read elsewhere as a starting point to test against your own numbers, not a rule to chase blindly.
How to Use Creative ROAS to Decide What to Produce Next
Once you have a reliable creative-level ROAS for several ads, use it to steer production, not just to grade past work:
- Scale winners by variation, not duplication. If a hook is working, produce new openings around the same core angle rather than copying the exact ad into new ad sets, which accelerates fatigue.
- Kill fast, but only after the threshold. If an ad is below break-even after your minimum spend and time window, pause it and move budget to the next test.
- Feed the brief loop back into production. Whatever is working — a specific proof point, pacing, or first frame — should go straight into your next ad creative brief, so the pattern gets repeated on purpose instead of by accident.
- Keep a steady supply of fresh tests. Creative ROAS is only useful if you have enough new creative in rotation to replace what's declining — which is where a creative testing calendar helps keep the pipeline moving instead of stalling once a winner fatigues.
How FrameNotion Fits Into Creative ROAS Testing
Calculating ROAS on video ad creative only matters if you have enough creative variety to actually test — a single winning video can't tell you much about what made it win. FrameNotion is built for that volume problem: paste a product link, and FrameNotion AI writes and renders a 30-second vertical ad (hook, problem, benefit, proof, offer, call to action) in about 10–20 minutes, ready for TikTok, Reels, Shorts, or Meta ads.
Because each ad is generated from scratch rather than from a template, you can brief several distinct hooks on the same product, launch them side by side, and measure which one actually earns its ROAS once real spend starts flowing through it. After the first test results come in, text and color changes can be made directly on the same ad without starting over, and Pro and Agency plans can generate A/B hook variants to feed your next round of testing. If you're new to the process, FrameNotion's examples page shows finished ads across different products and styles before you commit any budget to them.
One honest limitation worth stating plainly: FrameNotion doesn't publish ads directly to ad platforms or report back on how they perform — you still launch the ad yourself and pull spend and revenue numbers from Ads Manager using the method above.
The first ad is free on sign-up, so you can test the creative-ROAS workflow on a real product before deciding whether to commit to a paid plan.
Frequently asked questions
Is creative-level ROAS the same as blended account ROAS?+
No. Blended ROAS averages revenue and spend across every ad in your account, which hides whether any single video is actually profitable. Creative-level ROAS isolates revenue and spend for one specific ad so you can judge it on its own.
How long should I wait before judging a new video ad's ROAS?+
There's no universal number, but a reasonable starting rule to test is waiting until the ad has spent roughly three times your average order value, or run for at least 48-72 hours, whichever comes later. Judging earlier usually means reacting to noise.
Why does my creative ROAS look different in Ads Manager than in my own spreadsheet?+
This usually comes down to attribution window mismatches, lifetime versus rolling-window reporting, or audience overlap between creatives in the same ad set. Standardize your attribution setting and time window before comparing numbers across ads.
Should I calculate ROAS the same way for a video ad as for a static image ad?+
Yes, the formula doesn't change by format. Pull spend and attributed revenue for that specific ad over a consistent window and divide revenue by spend, whether it's a video, carousel, or static image.
What ROAS should I aim for on a new video ad?+
There's no fixed good number — it depends on your margin and average order value. Calculate your own break-even ROAS from cost of goods, fees, and shipping first, then treat any published benchmark as a starting point to test, not a target to chase blindly.