META ADS CASE STUDY
Scaling an e-commerce account to $525K in campaign revenue while running promotions at 4× ROAS
Online educational e-commerce platform · Meta (Facebook & Instagram) · Prospecting + seasonal promotions + segmented evergreen · 106-day period
The client
An online educational e-commerce platform selling digital and physical learning resources to a large, engaged consumer audience. The account runs a broad Meta program: an always-on prospecting campaign against the core catalogue, a rotating calendar of seasonal promotions aimed at warm and existing-customer audiences, and a set of segmented evergreen campaigns built around distinct audience interests. The brief was growth — put more revenue and volume through Meta while keeping campaign-level returns healthy.
The challenge
- A single campaign carried the account. One flagship prospecting campaign drove the majority of revenue, so scaling meant pushing its spend up several-fold without letting returns collapse.
- Promotions had to convert fast. Seasonal sale events needed to be built and timed to move warm audiences efficiently inside short windows.
- A sprawling structure hid the spread. A wide set of segmented campaigns had to be read individually — funding the ones earning their keep, not over-investing in the ones that weren’t — rather than judged on a blended average.
The approach
- Concentrated budget behind the winner. Scaled the core prospecting campaign roughly 5× in spend while it stayed the account’s primary revenue engine — accepting a deliberate efficiency-for-volume trade rather than protecting an unsustainable return at low spend.
- Built and timed seasonal promotions. Ran short-window sale events to warm and existing-customer audiences — the account’s highest-returning activity, at roughly 4× ROAS.
- Kept the structure segmented and legible. Maintained distinct audience and interest campaigns so each could be funded on its own performance instead of a blended number that masks the spread.
- Reviewed on a disciplined cadence. Read recent movement against a longer baseline before acting, so nothing got called a trend on noise.
The results
Over the latest 106-day window, the flagship prospecting campaign generated $525K in tracked revenue from 6,291 purchases at a 3.0× return. Scaling its spend roughly 5× from the prior window nearly tripled its revenue; the return compressed from the prior window’s ~5.4× to a still-strong 3.0× — an intentional trade of efficiency for volume, not a slip. Alongside it, the seasonal promotion event returned ~4.0× blended on ~$28K of spend, and account-wide average order value rose roughly 50%.
| Featured campaign | Spend | Revenue | Purchases | ROAS |
|---|---|---|---|---|
| Flagship prospecting — core catalogue | $174,500 | $525,299 | 6,291 | 3.01× |
| Seasonal promotion — warm audiences | $11,097 | $47,280 | 466 | 4.26× |
| Seasonal promotion — core segment | $10,896 | $44,363 | 446 | 4.07× |
| Segmented evergreen — niche audience | $5,047 | $14,638 | 135 | 2.90× |
| Featured subtotal (4 of the account’s campaigns) | $201,539 | $631,579 | 7,338 | 3.13× |
Featured campaigns are a subset of the account’s sales program; the account’s full sales-objective totals for the window are visible in the screenshot below.
Why it worked
The growth came from concentration and timing, not scattergun spend: backing the one prospecting campaign that was already proving itself and scaling it hard, layering high-intent promotions onto warm audiences at the right moments, and keeping the segmented structure legible enough to fund by performance. Scaling always costs some efficiency — the judgement was in trading it deliberately, nearly tripling the flagship’s revenue while keeping the return strong, rather than letting it erode by accident.
Scope: figures are campaign-level Meta results for the featured campaigns over the stated window, selected to illustrate scaling and promotion performance; they are not a statement of total blended account ROAS. Brand identity, product categories, campaign names, and account owner have been anonymized.