By Alexander Huba | Originally published on Towards Quantum
What to decide about quantum computing this year, and what to ignore. Most executives are thinking about quantum the wrong way — not because they don't understand the physics, but because they start with the technology and then go hunting for a use.
Quantum computing has entered its keynote era.
The headlines say "breakthrough." The vendors say "transformational." Your board has started asking about the plan. And somewhere in your company, someone is building a slide with a glowing blue sphere on it.
Here's my hot take: most executives are thinking about quantum the wrong way. Not because they don't understand the physics. You don't need to. They're getting it wrong because they start with the technology and then go hunting for a use.
Asking "what's our quantum strategy?" is like asking "what's our scalpel strategy?" If you're not doing surgery, you don't need one. If you are, you'd better know exactly where to cut.
Tools don't get strategies. Decisions do.
So what is a quantum computer actually doing for you?
One analogy, then we move on.
A regular computer thinks in bits. Each bit is a coin lying flat on the table: heads or tails. Definite.
A qubit is a coin still spinning in the air. It hasn't committed yet. A quantum algorithm choreographs a whole table of spinning coins. By the time they land, the wrong answers have cancelled each other out and the good ones have gotten louder.
Notice what I didn't say. I didn't say it "tries every answer at once." That's the myth that sells conference tickets. If it were true, quantum would solve everything. It doesn't.
The choreography only works on problems with a very particular shape. So quantum isn't a faster laptop. It's a fussy specialist. Your job isn't to understand the coins. It's to recognize when you have a problem worth calling the specialist for.
The three buckets
Almost every pitch you'll hear lands in one of three buckets. They are not equally believable.
Optimization: the one everyone's selling
Picture a logistics company routing 500 trucks through 3,000 stops. The combinations are astronomical, so it sounds like a perfect quantum job.
Tough love: the known quantum speedups for problems like this are modest, and classical solvers are already very good. Besides, plenty of bad routes come from bad data, not weak math. If half your addresses are wrong, a quantum computer will give you a gorgeous answer to the wrong question. Fix the data first.
Simulation: the one with the best excuse
Molecules and materials follow quantum rules, so a quantum computer can model them in their native language. Think of a battery maker screening new materials, or a pharma team deciding which drug candidates deserve lab time.
The realistic win isn't "quantum discovers a miracle drug." It's fewer doomed experiments, most likely on specific hard problems rather than all of chemistry at once. That's not a headline. It's a budget line.
Machine learning: the one with the most asterisks
The pitch: quantum will spot fraud patterns your current models miss.
The reality: getting large amounts of data into a quantum computer is a hard, unsolved problem. And several celebrated quantum machine learning speedups have since been matched by clever classical algorithms. If a vendor leads with this, ask for one case where it beat your best model on your data. Then enjoy the silence.
The part the keynotes skip: timing
Today's quantum machines are real, improving, and still noisy. Even the most credible recent "advantage" demos have tackled physics problems, not business ones. The machines that could deliver broad commercial value need serious error correction. IBM and Google are both aiming for 2029 for their first large-scale versions.
Treat those as targets, not delivery dates. Anyone who gives you a confident year for when quantum pays off for your business is selling something. Possibly a conference ticket.
So no, you don't need to buy a quantum computer. Most companies never will. When it's worth it, you'll rent time through the cloud.
What to do this year
But ignoring quantum entirely is a different mistake. Here's what to do this year.
Upskill a few translators. Not a physics department. Two or three people who can sit between a vendor and a business owner and ask, "Impressive. What does it change for us?"
Identify your bottlenecks, not your use cases. List your most expensive computational headaches: the simulation that takes weeks, the optimization where a 2% gain would be worth millions. Check whether any of them have the right shape. Most won't. That's a cheap, respectable answer.
Monitor the ecosystem. Track what competitors actually ship, not what they announce. When an "advantage" headline lands, ask: advantage over what, on which problem, and does it look anything like mine? Twice a year is plenty.
One more thing, and it isn't optional
Ask your CISO about post-quantum encryption. A future quantum computer could break the encryption protecting today's data, and anything stolen now can be decrypted later. The U.S. government has already set 2030 and 2031 deadlines for its most sensitive systems. That clock is running whether or not you ever touch a qubit.
Start with the decision
Every emerging technology runs the same play. The internet. Cloud. AI. Now quantum.
The leaders who win aren't the ones who jumped in first to look innovative. They aren't the ones who ignored it until it bit them, either. They're the ones who kept asking the least glamorous question in the building:
What are we trying to decide, and is this the right tool for it?
FOMO is not a strategy. Neither is denial.
Start with the decision. The qubits can wait in the lobby.
