Unable to load audio. Please try again later.
In a market where specialists rarely change jobs, the surest way to hire a team is to buy the firm.
Outside China there are little more than 5,000 quantum researchers, and five Big Tech companies already employ nearly a quarter of them, up from one in sixteen in 2015. Once inside, they stay put: 73 per cent of the experts who joined Google, IBM, Nvidia, Microsoft or Amazon are still at the same company, and those who do move mostly go to another giant.
That makes quantum an unusual labour market. In most fields, a rival with enough money can hire the people it wants. Here, Zeki Data’s mapping of the five companies’ hiring shows that they mostly do not try: 41.9 per cent of their quantum hires came straight from universities and only 8.6 per cent from quantum companies. Among quantum companies, only IonQ has persuaded meaningful numbers of their people to leave.
Quantum talent goes in from universities, and then stays put
Share of hires into Google, IBM, Nvidia, Microsoft and Amazon, by sector of previous employer
Source: Zeki, Phasecraft Diagnostic, October 2026
The deal book is beginning to show it: if you can’t hire a team, you buy one. Intel has agreed to buy Classiq, a quantum software company that has only raised about $200 million, for a reported $2.3 billion, which CTech called the most expensive quantum acquisition to date by a significant margin. IonQ paid about $1 billion for Oxford Ionics, and Google bought Atlantic Quantum. Quantum Computing Report argued that talent, as well as technology, lay behind those deals.
The useful question about any quantum company is therefore what a buyer would see. Phasecraft, a UK algorithms company that grew out of UCL and the University of Bristol, is a good test. It writes algorithms for quantum computers that other people build, has raised more than $50 million including grants, and works with Google Quantum AI, IBM, Quantinuum and QuEra, according to its Series B announcement. Consider its workforce in three parts.
The first part is scarcity. Britain’s market for quantum algorithm experts is tiny, and two companies dominate it. Quantinuum holds 36.1 per cent of the experts at the country’s twelve biggest employers of them, and Phasecraft 26.5 per cent. Between them they hold nearly two-thirds. Anyone who wants a serious algorithms bench in the UK has few other places to look.
The second is quality. Phasecraft has 23 researchers in our comparison to Quantinuum’s 153, yet its median citations are 986 against 649, and 35 per cent of its researchers hold a prestige award against 16 per cent. Quantinuum has the edge on one measure, with a third of its researchers trained at top research institutes against 22 per cent. On most measures, though, a buyer would find the smaller team the stronger bench.
A team of 23 that outscores one of 153 on most measures
Phasecraft and Quantinuum researchers compared
Median citations: Phasecraft 986, Quantinuum 649. Shares are of researchers in the comparison, except the first row. Source: Zeki, Phasecraft Diagnostic, October 2026
The third is steadiness. Quantum companies are tempted to retreat towards classical computing, where the revenue is. QC Ware announced exactly such a pivot in April 2023, and over the following two years its research staff drained away, mostly to Nvidia, with some landing at Phasecraft. Phasecraft never swerved from quantum algorithms, and a buyer would find a team that has not been hollowed out.
The first doubt is commercial. Technical staff make up 90 per cent of Phasecraft’s workforce, against 74 per cent at QC Ware and 63 per cent at Qunasys. It launched its first software platform, Mondrian, in October 2025 and has since added one business development lead. None of its 17 open roles are in sales. That is a weakness for a company meaning to live on revenue. For a buyer who wants the researchers rather than the customers, it is almost a selling point.
Phasecraft is among the most technical teams in its peer group
Share of workforce in technical and go-to-market roles (%)
Source: Zeki, Phasecraft Diagnostic, October 2026
The second doubt is Washington. Phasecraft opened an office there in 2024, has joined DARPA’s Quantum Benchmarking Initiative and counts a former National Science Foundation programme director among its senior hires, and five of its open roles are in the capital. Washington holds 2.5 per cent of America’s quantum research experts, against 28.9 per cent in the San Francisco Bay Area, so as a recruiting base it is a puzzling choice. As a way to sit close to research funders it makes more sense, and it points to the same conclusion: this is a company organised around research money rather than customers.
Washington holds 2.5% of US quantum experts. The Bay Area holds 28.9%
Share of US-based quantum research experts, by city (%)
Phasecraft’s US base highlighted. Source: Zeki, Phasecraft Diagnostic, October 2026
The catch is neutrality. Part of Phasecraft’s value is that its algorithms run on Google, IBM, Quantinuum and QuEra machines alike, and a sale to any one of them would put the other relationships at risk. Intel recognised the same problem with Classiq, promising to keep its software compatible with other makers’ hardware. That complicates a deal with a hardware company. It does not remove the logic of one.
Three things would change the reading: Phasecraft building a sales team before it needs one, Big Tech starting to hire from quantum companies, or buyers finding another way to reach the same talent. Until one of them happens, the arithmetic holds. In most industries, a rival pays to hire the person. In quantum, it increasingly has to buy the firm that employs them.
About this analysis: drawn from the Zeki Diagnostic Report on Phasecraft, October 2026, built from publicly available profile data for Phasecraft and eight comparator quantum companies (Algorithmiq, Kvantify, Qubit Pharma, Qunasys, HQS Quantum Simulations, QC Ware, Haiqu and Entropica Labs), and for Google, IBM, Nvidia, Microsoft and Amazon. Researcher comparisons cover 23 Phasecraft and 153 Quantinuum researchers. Function measures are based on job titles. The full diagnostic also covers research hiring quality, retention and go-to-market hiring.
Citations: Bloomberg via KSL (Intel and Classiq); CTech (deal size record); Quantum Computing Report (acquisitions and talent); HPCwire (Phasecraft funding and partners).