Payments
Attach Growth Compresses Blended GM but Expands Operating Margin
At commerce platforms where an embedded payments segment is growing
faster than the software/subscription segment, blended gross margin
falls structurally — but operating margin can still expand. The
mechanism: payments processing is largely automated (marginal OpEx is
near-zero per transaction), so the R&D and S&M base built to
support the software segment extends across the payments volume at
negligible incremental cost. Revenue scales faster than OpEx even as the
GM mix dilutes.
This creates an analytical trap: a analyst watching blended GM
compress quarter after quarter may conclude the business is
deteriorating, while the correct read is that a higher-volume,
lower-margin but operationally-leveraged payments segment is diluting
the blend without impeding profit generation.
Evidence
- SHOP Q4 FY25: Merchant Solutions (payments-dominant) at 37% GM now
79% of revenue and growing faster than Subscription Solutions (81% GM).
Blended GM fell 2.0pp YoY to 46.1% — lowest in 12 quarters. Yet
operating margin hit a record 17.2% (+0.7pp YoY). OpEx grew 20% while
revenue grew 31%. Rule of 40: 49.6.
- The divergence is structural: every 1pp increase in GPV penetration
(currently 68% of GMV) generates incremental revenue at ~37% GM with
near-zero incremental fixed cost — pure operating leverage at the bottom
of the stack.
- Payments mix is expected to continue rising (runway to 80%+),
meaning blended GM compression will continue even as operating margin
expands further.
Implication
For any platform undergoing payments-attach growth (Shopify, Toast,
Block, WEX, PayFac-embedded SaaS), do not use blended gross margin as
the quality screen. Use three alternative checks:
- Operating margin trajectory — the true leverage
signal; if expanding despite GM compression, the model is working.
- Gross profit dollar growth rate — if growing ≥
revenue growth rate, the mix shift is not destructive at the
profit-dollar level.
- Payments penetration rate direction — rising
penetration at ~37% GM is mechanically dilutive to blended GM and should
be modeled as a structural headwind, not an anomaly. Build a 2-3pp
annual GM compression assumption into forward FCF models for platforms
still in payments ramp phase.