Why this matters before the hardware arrives
Quantum computing is at the stage classical computing was in the early 1950s: real machines exist, they do genuinely useful things in narrow domains, and almost every claim about general-purpose advantage is premature. That is not an argument for ignoring it. It is an argument for being precise about which parts of your planning horizon it touches.
There are two questions worth putting on a board agenda now. The first is cryptographic: what data are you holding today that must still be confidential in 2035? The second is operational: do you have a class of optimisation problem that is currently solved badly because solving it well is intractable?
The cryptography question is the urgent one
Harvest-now-decrypt-later is not speculative. An adversary who captures encrypted traffic today can store it until a machine capable of breaking the key exists. For most operational data that is irrelevant — a session token from 2026 is worthless in 2035. For medical records, financial positions, legal correspondence and identity documents, it is not.
NIST's post-quantum standards are published and implementations are shipping in mainstream TLS libraries. The practical work in 2026 is inventory, not migration: knowing where your long-lived secrets are, what algorithms protect them, and which of your vendors can tell you the same. Organisations that have done that inventory will migrate in months. Organisations that have not will spend that time finding out what they have.
The optimisation question is the interesting one
Portfolio optimisation, route planning under many constraints, molecular simulation and certain scheduling problems all have structure that quantum algorithms may eventually exploit. The honest position today is that classical heuristics still win on nearly every real instance, and the crossover point is unknown.
What is worth doing now is framing the problem properly. Teams that have expressed a scheduling problem as a clean quadratic optimisation are in a position to test quantum and quantum-inspired approaches cheaply when they mature. They also, almost always, discover that better classical solvers improve their result substantially in the meantime. The framing pays for itself either way.
What we would advise
Assign someone to own cryptographic inventory this year. Include post-quantum readiness in vendor assessments. Pick one optimisation problem where you suspect you are leaving money on the table and formalise it — then solve it classically and keep the formulation. Ignore everything else until the hardware gives you a reason not to.