MER is computed as total revenue / total marketing spend and reported as a unitless multiple: $1M in revenue against $250K in marketing spend produces a MER of 4.0. The “over a defined period” qualifier matters — MER is a window metric, typically measured weekly or monthly, and the window has to be long enough to absorb the lag between spend and revenue.
The structural difference from ROAS sits in both the numerator and the denominator. The numerator is every order in the period — paid, organic, direct, email, referral — not just what platforms attribute. The denominator is total marketing spend across every channel: paid social and search, affiliate and influencer payouts, retention email and SMS tooling, agency fees, creative production. It overlaps with the acquisition-spend numerator of CAC but is broader, because it also covers retention.
MER rose to prominence post-iOS 14.5 as platform-reported ROAS grew noisier and operators needed a measurement that sidestepped attribution. The trade-off is bluntness: MER tells you whether marketing in aggregate is working but cannot tell you which channel to cut. Pair it with per-channel ROAS for allocation, and watch the organic/paid mix — a MER that improves because organic grew is a different story from one that improves because paid efficiency did.