Case Study · E-commerce / Paid Media
A high-AOV e-commerce account where Meta's platform reporting was treated as directional rather than gospel, Triple Whale linear attribution drove every real decision, and paid media best practice was applied steadily across 24 months.
$475k → $865k
TW-attributed revenue, year on year — up 82% in Triple Whale linear attribution
+82%
TW revenue growth (YoY) $475k → $865k
14.3×
Meta-reported ROAS (28d-click) Up from 9.3× — same windows
$1.34M
Total TW-attributed revenue Across two-year engagement
$94k
Peak monthly TW revenue July 2025
The Challenge
The client is a high-AOV e-commerce brand. By every in-platform metric, the Meta account looked fine: Meta's 28-day-click ROAS sat comfortably between 8× and 11× month after month. The problem wasn't a broken dashboard. The problem was that Meta — like every ad platform — over-reports its own contribution, and the team had no trustworthy way of pressure-testing those headline numbers against the P&L.
For this account the over-reporting ratio sat at around 12× in the opening period, measured against Triple Whale's linear attribution view of Meta's true contribution. That kind of gap makes confident decision-making impossible: scaling spend, killing underperformers, validating creative tests — every call hinges on a number you can either trust or you can't.
The engagement wasn't framed as a turnaround. It was set up as a two-year programme of paid media best practice — measurement first, then account architecture, then creative and audience — with Triple Whale linear attribution as the navigation tool and Meta's own reporting kept on hand as supporting context.
The Approach
Three workstreams running across the full two years, sequenced so each one made the next more effective. Triple Whale linear attribution guided every spend and creative decision; Meta's own dashboard was kept as supporting context, not the source of truth.
Phase One
Triple Whale was implemented with linear attribution adopted as the navigation tool for every spend decision. Conversions API was deployed to feed cleaner first-party signal back into Meta's algorithm. Weekly TW reviews became the cadence for budget and campaign calls. Meta's own dashboard stayed visible as a check on platform-reported volume — useful for context, never the deciding number.
MONTHS 1–2
Phase Two
A fragmented campaign structure was consolidated into a cleaner TOF / MOF / BOF model — fewer, larger campaigns to give Meta's algorithm more learning signal. An Advantage+ Shopping (ASC) layer was launched in May 2025 organised around top product categories, and quickly became the second-largest spend contributor. Underperformers identified through the TW efficiency view were cut, with budget re-concentrated into the proven mid-funnel conversion campaign and bottom-funnel dynamic retargeting.
MONTHS 3–12
Phase Three
A quarterly creative refresh cadence was established to stay ahead of fatigue, with a multi-format production pipeline covering UGC, static, motion and carousel across all key aspect ratios. Catalogue and feed work tightened dynamic retargeting. Audience strategy moved from narrow interest stacks to broad-with-creative-as-targeting, with lookalike seeds rebuilt from high-LTV purchaser data. Creative judgement calls were made on TW lift, not in-platform engagement signals.
MONTHS 6–24
The Results
Both measurement views moved together: TW-attributed revenue grew 82% year on year (linear attribution), while Meta's own 28-day-click ROAS lifted 54% across comparable windows. The fact that the truer view grew faster — with the over-reporting gap narrowing from 11.8× to 9.2× — is the signature of real incremental improvement rather than optimisation against the platform pixel. Revenue from the brand's Triple Whale workspace (linear attribution); Meta-reported ROAS from Windsor.ai using a 28-day-click attribution window.
The Trajectory
Monthly TW-attributed revenue across the engagement — Triple Whale linear attribution. From a $21k floor in June 2024 to a $94k peak in July 2025, with the long-run monthly average lifting from $39.6k in Year 1 to $72.1k in Year 2.
Work With Me
If you're running Meta (or Google, or TikTok) and want a steady hand on the account — trustworthy measurement, disciplined best practice, and consistent month-on-month improvement rather than a hero quarter — let's talk.
Back to all case studies