In any company with both hardware/product and software/subscription revenue, blended gross margin compresses when the lower-margin segment grows faster than the higher-margin segment. This works in both directions: (1) SaaS-primary companies adding AI hardware see hardware drag the blended GM down; (2) hardware-primary companies building a subscription layer see GM compress when hardware demand surges faster than subscription growth. The critical distinction: whether the compression is recoverable (pricing power exists, subscription will catch up) or structural (hardware permanently dominates mix). Operating margin can simultaneously expand through operating leverage even during GM compression — track both, not just GM.
When screening for gross margins in hybrid hardware/software companies: (1) always compute segment-level margins separately — blended GM misleads in both directions; (2) when GM compresses, identify which segment is outgrowing the other and whether it is demand-driven (cyclical risk) or AI-structural (secular tailwind with temporary mix effect); (3) check whether the company has validated pricing power (NPS, customer lock-in, no easy substitutes) — if yes, GM compression from mix-shift is recoverable via pricing action, not a structural deterioration signal; (4) look at op margin alongside GM — if op margin is expanding or stable while GM compresses, the business is not actually getting weaker, just reflecting a different revenue mix; (5) for hardware-primary hybrids, the recovery cycle is: AI demand surge → hardware accelerates faster than subscription → GM compresses → pricing action flows 1-2 quarters later → GM recovers → subscription catch-up reinforces. Flag this sequence rather than treating the initial compression as bearish.