Step 08 of 12 · 20 min
Project frame and classical baseline
Choose one of the four project types and write the classical baseline before any quantum run. A hardware run without that measurement is a demonstration. The handbook section is the long form.
What you are actually going to build
Quantum advantage means demonstrating, through a fair comparison, that a quantum method outperforms the best relevant classical approach on a useful problem. You are not expected to prove that during a 48-hour hackathon. Your goal is to build a well-scoped project, create credible evidence, and explain honestly what the evidence does and does not show.
A mapped problem
Start from a real industry problem. Name who has it and how it is handled today, then write the variables, constraints, objectives, data, and success metrics. Map that statement to a QUBO, a Hamiltonian, a kernel, a circuit, or another quantum representation. A small simulator or approved-hardware example is useful when it stays honest about size. State the gap between the demo and a meaningful instance, and name what must change in the hardware, the algorithms, the data, or the industry conditions. A Domain/Industry Expert or other subject-matter expert is especially valuable here.
Mapped problem in the handbook
Assess an optimization problem with the Next-Step Quantum Decision Guide (opens in a new tab)
A benchmark or comparison
Hold one bounded problem fixed and run it classically and with the quantum method. Use the same definition and comparable inputs. Choose the metric before the run, then record configurations, instance sizes, timing, quality, and resource use. You can discuss where a future crossover might occur and what blocks it. A project does not have to find or predict a crossover. Report the result when the classical method stays ahead.
Benchmark or comparison in the handbook
Open the benchmarking and readiness guide
Next-Step Quantum Decision Guide for an optimization comparison (opens in a new tab)
A tool
A tool might visualize a circuit or a state, translate a domain format into a QUBO, a Hamiltonian, a kernel, or a circuit, teach a step, assist a workflow or a benchmark, or track resources and evidence. Say who it is for, what they provide, what they receive, why that matters, and where it stops. A polished interface without a meaningful user problem is not enough.
An analysis
Assess whether quantum computing fits a sector or a problem. Compare the classical methods used now with quantum approaches that might apply, including the hardware and resource gap. Policy, risk, deployment, and readiness are in scope. A well-supported “not a fit yet” or “not quantum-shaped” conclusion is a complete project when the technical claims match current hardware and algorithms, and the sources are credible.
Who leads which type
Mapped-problem and analysis projects can be led primarily by Domain/Industry participants. Benchmark and tool projects generally require more Builder/Developer work. All four become stronger when someone understands what the output means in the real world.
Choose your next step
The decision guide is for optimization-shaped problems. Do not force simulation, chemistry, classification, link prediction, or other non-optimization problems through an optimization-only assessment. Those projects still use: Problem → Current method → Classical baseline → Quantum hypothesis → Experiment → Metrics → Evidence → Limitations → Recommendation.