Quantum software spans programming languages, circuit synthesis, compilation, resource estimation, error mitigation, error correction and domain applications. These layers should not be valued as one category: a hardware-independent workflow tool has a different market from a compiler optimised for one processor, while an application company must prove that quantum resources improve a customer outcome rather than merely reproduce a research demonstration.
The transition towards fault tolerance changes software requirements. Resource estimation must connect algorithms to physical error rates and code overhead; compilers must manage architecture-specific connectivity and timing; decoders have to process error information within strict latency constraints. Classical high-performance computing remains part of the workflow for optimisation, simulation, verification and orchestration. The most defensible products may therefore sit at the interface between quantum and established computing rather than inside an isolated quantum stack.
Commercial evidence includes active usage, repeatable deployment, hardware portability, measurable reduction in required resources and integration with existing scientific or enterprise tools. Partnerships and open-source adoption can expand reach but do not by themselves establish monetisation. QFM classifies companies by the bottleneck their software removes and links product claims back to published documentation and customer evidence.