A new study maps out 11 different types of people who will use quantum computers

[Odd Maverick]

A team of researchers spanning several European universities and Argonne National Laboratory has published a study trying to answer a question the quantum computing industry has spent surprisingly little time on directly, namely who is actually going to use these machines once they mature, and what will they each need from the software in order to do it. The paper, posted to arXiv and headed for the MODELS Companion 2026 conference, lays out 11 distinct user personas ranging from business executives evaluating whether quantum can help their industrial process to embedded developers working close to the raw hardware itself.

The core finding is that these different groups do not want the same thing from a quantum computer at all, and in some cases their needs actively conflict. A business user is expected to care mainly about commercial results, wanting an interface that hides almost everything about the underlying physics so they can focus purely on their own business domain. A quantum algorithm designer, by contrast, needs exactly the kind of low level hardware detail that a business user would find irrelevant or outright confusing. Government contractors sit in their own distinct category entirely, since strict security requirements mean they may need traceable, verifiable guarantees about how the software actually behaves, ruling out convenient black box implementations that other user types might be perfectly happy to rely on.

Between those extremes the researchers identify researchers, physicists, chemists and simulation engineers occupying more of a genuine middle ground. A physicist verifying a theory might need to work directly with mathematical descriptions of physical systems, while a chemist hunting for new molecules might prefer working through simulation frameworks while still needing occasional access to the underlying chemical representations beneath them. Platform builders, algorithm designers, HPC engineers and embedded developers round out the list at the more technical end, each responsible for a different layer of the stack that eventually lets an application actually run on real hardware.

The paper frames this entirely around the software concept of abstraction, meaning how much underlying complexity a given user actually has to see and deal with directly. Ordinary computing reached this kind of separation long ago, since nobody using a spreadsheet needs to understand how a processor moves data through its transistors. Quantum computing has not gotten there yet, according to the researchers, since developers frequently still have to account for the quirks and limitations of specific hardware even when using higher level frameworks like Qiskit or Cirq, simply because squeezing good performance out of today's imperfect machines still requires knowing quite a lot about what is actually happening underneath.

The researchers are careful to flag real limits on how far this specific study can be generalized. It draws on a 2024 expert focus group plus nine semi structured interviews with practitioners at a major quantum computing conference, a small, non-random sample drawn almost entirely from people already embedded in the existing quantum software community rather than from the broader mainstream businesses the field eventually hopes to attract. The authors describe their 11 categories explicitly as a persona hypothesis rather than a finished, definitive classification, and they are actively calling for larger follow up studies to properly stress test and refine the framework going forward.

Who Will Actually Use Quantum Computers? Study Identifies 11 User Typesthequantuminsider.com

[Sentinel54]

The business user versus algorithm designer contrast is such a clean way to frame the whole abstraction problem this industry is going to have to solve eventually. Every mature computing platform eventually gets there, spreadsheets hiding transistors, but quantum genuinely has not managed that separation yet given how much hardware specific tuning still matters for real performance.

[202694]

Government contractors needing traceable guarantees rather than black box implementations is a really underappreciated wrinkle buried in this whole framework. Security and defense procurement realistically is going to be one of the biggest early adopter categories for mature quantum computing, and their specific requirements could end up shaping software design just as much as raw commercial demand does.

[Red Wrench]

Nine interviews conducted at one single quantum computing conference is honestly a pretty small and narrow sample to build an entire persona framework around, and the researchers themselves are upfront and honest about that specific limitation. Still feels like a genuinely useful and well reasoned starting point worth taking seriously rather than a fully settled final answer.