Articles / For researchers

Leaving academia for industry

What share of PhDs go to industry in ten fields, what they earn against postdocs, and which job-search steps people who left academia found useful.

Many researchers reach the middle of a PhD knowing they are good at research and that they do not want an academic career. They then find that nobody around them can describe the alternative. Supervisors know the academic route well because they took it. University careers offices see few PhD students.

So most people plan the move from a handful of stories about someone who left. This post sets out what the public data shows instead. Most of it comes from the US National Science Foundation and the UK organisation Vitae, with surveys of PhD graduates and of the managers who hire them.

It covers six things:

  1. What is normal in ten fields
  2. Why your own situation may differ
  3. Whether to do a postdoc
  4. Who to ask for advice
  5. How to look for a job
  6. What to do, and when

1. What is normal in ten fields

The table has two figures for each field. The first is the share of new PhDs who go into a postdoc. The second is the share of the rest, those who take a job, who go to industry.

FieldGo into a postdocOf those taking a job, go to industryIndustry payPostdoc pay
Most go to industry
Engineering36%74%$125,000$65,000
Physical sciences55%71%$120,000$65,000
Computer and information sciences25%65%$180,000$70,000
About half go to industry
Biological and biomedical sciences65%52%$120,000$60,000
Mathematics and statistics43%50%$150,000$70,000
A minority go to industry
Agricultural sciences48%39%$100,000$60,000
Geosciences61%37%$110,000$65,000
Psychology56%32%$97,000$58,000
Health sciences39%26%$120,000$61,934
Social sciences29%19%$129,000$65,000

US doctorate recipients, 2024. Postdoc column: share of those with firm plans who were heading to a postdoc. Industry column: share of those with a non-postdoc job who were going to industry or business. Pay is the expected median salary. Sources: NSF Survey of Earned Doctorates 2024 and NSF 26-312.

Engineering, physical sciences and computing

In these fields, between two thirds and three quarters of PhDs who take a job go to a company. Employers have hired PhDs from these fields for decades. The roles are established, and you will not have to explain why someone with a doctorate is applying.

Biology and mathematics

These fields changed most over twenty years. In 2004, 28% of biology PhDs who took a job went to industry. In 2024 it was 52%. In mathematics the share rose from 19% to 50%.

The biology figure needs a qualification. About two thirds of biology PhDs with firm plans go into a postdoc, the highest rate of any field. The 52% describes the remaining third. Most biologists who reach industry do so after a postdoc, and section 3 covers what that costs.

Agriculture, geosciences, psychology, health and social sciences

Fewer than four in ten PhDs who take a job in these fields go to industry. In the social sciences it is fewer than two in ten. Many of the others go to government and nonprofits.

Those who do go to industry are paid well. Social science PhDs expect a median of $129,000, slightly more than engineers. The roles are often adjacent to the discipline, such as user research, data science, policy and evaluation, so you will need to describe your methods in terms the employer uses.

Outside the US

Other countries with tracking data show a similar picture. In the UK, 41% of 2022/23 doctoral graduates in work were employed in higher education fifteen months later. In a sample of 1,432 graduates from eight European universities, 40% worked in academia and 39% in business. Of University of Toronto graduates from 2000 to 2015, about half worked in post-secondary education and 26% held tenure-track posts.

Sources: Vitae 2026; Sánchez-Barrioluengo, Lawson and Cañibano 2026; Reithmeier et al. 2019.

2. Why your own situation may differ

  1. Industry hiring has slowed. The NSF reports that industry's share of new PhD jobs has fallen in every science and engineering field since at least 2022. A search now is likely to take longer than it did in 2021.
  2. The figures are first destinations. Many postdocs move to industry later and are not counted in the industry column.
  3. The salaries are what graduates expected to earn in the job they had accepted, and they are US figures.
  4. A subfield can differ a lot from its field. The market for a computational biologist is closer to computing than to ecology.

3. Whether to do a postdoc

If you intend to work in industry, a postdoc costs you money. Postdoc pay is similar in every field, between $58,000 and $70,000. Industry pay ranges from $97,000 to $180,000. In biology, industry pays twice the postdoc salary.

The difference persists. The economists Shulamit Kahn and Donna Ginther followed 10,402 US biomedical PhDs. Ten years after graduating, those who had done a postdoc earned 11% less than those who had gone straight into work. Among those who ended up in industry, the cumulative shortfall was about $240,000. The authors concluded that employers outside academia do not pay more for postdoc training.

That study covers people who graduated between 1980 and 2010, in one field and one country, and some industry research roles do require postdoctoral experience. A reasonable approach is to do a postdoc when a role you want requires it, and not as a way to delay choosing.

Source: Kahn and Ginther, Nature Biotechnology 2017, as reported in Nature.

4. Who to ask for advice

Most supervisors cannot advise on industry careers, because they have not had one. The surveys show this clearly. Among UK PhD students, 82% say their programme prepares them well for a research career and 40% say it prepares them for a non-research career. Only 27% say their supervisor has useful advice on careers outside academia.

A survey of 176 STEM PhD students and alumni, mostly in the US, found that 34% had their supervisor as their only professional mentor and 23% had no mentor. Only 18% had a mentor outside academia.

The most useful person to find is someone who left your own lab or department for industry in the last few years. They know your training and your supervisor, and they know what their employer looks for.

Sources: HEPI 2026, using Nature's 2025 PhD survey; Ganapati and Ritchie 2021.

5. How to look for a job

A University of Toronto survey asked 446 life-science trainees and alumni which career activities they had done. Those now working outside academia were also asked which had helped.

Networking events and career seminars were the most common activities. More than half had attended them, but only 22% and 15% rated them highly. Internships were rated the most valuable, and only a minority had done one. Informational interviews showed the largest difference between the groups: 53% of those who had moved to industry had done them, compared with 20% of current trainees.

A separate US survey asked 56 STEM alumni how they found their first job. The sample is small. Just over half found it through a person.

How they found their first jobRouteShare
Cold applicationCold application30%
Supervisor or facultyThrough a person27%
Own networkThrough a person21%
PhD peersThrough a person9%
Career fair or centreThrough a posting or event5%

Source: Ganapati and Ritchie 2021. The remaining 7% gave other answers.

Nature surveyed 1,134 scientists who hire for their labs, 381 of them in industry. Three of the findings are relevant to applicants.

  • 73% of industry hirers rated the quality of their candidates as high. They said they look for relevant knowledge, fit with the team and the ability to explain your own science.
  • 61% of industry hirers reply to unsolicited applications. They are most likely to reply when they have a vacancy (55%) or when the applicant knows the team's research (48%).
  • 63% of industry hirers say candidates frequently negotiate pay. In academia the figure is 28%.

Sources: University of Toronto survey, bioRxiv; Nature Careers, Hiring in Science, fieldwork June 2024.

6. What to do, and when

The actions below follow from the evidence above. The timing is our suggestion.

A year or more before you finish

Contact two or three people who left your lab or department for industry. Ask your supervisor and your funder whether you can do a placement. In the UK, between 33% and 37% of doctoral researchers do one.

Six to twelve months before

Ask about ten people in roles you are considering for a 20-minute conversation about their work. Useful questions include:

  • What did your first three months involve?
  • What do new PhD hires on your team find difficult?
  • What did you have to learn that the PhD did not teach you?
  • How does your team decide who to interview?
  • Who else should I talk to?

Do not ask whether they have a vacancy. If they have one and think you are suitable, they will mention it.

A first message can be short:

Hi [name], I am finishing a PhD in [field] at [university], working on [one line]. I read your team's [paper, talk or product note] on [topic] and wanted to ask [a question about it]. I am looking at [type of role] and would be grateful for 20 minutes to hear how your team works. I can fit around your schedule.

The question about their work matters most, because hirers reply more often to applicants who know their research.

When you apply

Apply to advertised roles and also write to hiring managers directly. In the alumni survey, cold applications were the most common single route to a first job. On your CV, include examples of working with other people, such as a collaboration you led or someone you trained. A review of 11 studies found that employers value teamwork more than PhD students expect.

When you receive an offer

Negotiate. Use the industry median for your field as a reference and adjust it for your country.

What the data does not cover

  • Most countries. The tracking data comes from the US, the UK, Europe and Canada. If you trained elsewhere, check where ten alumni of your department now work.
  • How long a search takes. We found no reliable figures for the time from first application to offer.
  • Large samples for some questions. The alumni survey and the Toronto survey are small and self-selected.

Summary

  • Most PhDs who take a job go to industry in engineering, the physical sciences and computing. About half do in biology and mathematics, and a minority in other fields.
  • A postdoc reduces earnings for people who later work in industry.
  • Few supervisors can advise on industry careers. Former members of your lab who work in industry can.
  • Internships and one-to-one conversations were rated more useful than events.
  • Industry hiring managers reply to messages that show knowledge of their work, and they expect candidates to negotiate.

If you have better data for your field or country, tell us and we will update this post.

For researchers

Find roles that fit your research

Register once with your ORCID iD or Google Scholar profile. Asriel reads your publication record and gets in touch when a role fits. There is no application to write.