Running an Academic Lab is Not Running a Small Business

20 min read

Abstract

Academic principal investigators are often described as running 'small businesses' because they manage budgets, hire staff, and sell a vision. As someone who runs an actual small business, I find the analogy superficially appealing but structurally misleading. This post lays out where the comparison breaks down: who bears risk, who the customer is, what failure costs, and how time horizons differ. It is not a takedown of academia or of entrepreneurship, both of which are demanding in their own right, but an argument for retiring a comparison that flatters neither side when examined closely.

academiaentrepreneurshipsmall-businessresearch-managementincentives

Introduction

The analogy that running an academic STEM lab is similar to running a small business appeals to many people. Deans, funding agencies, PIs themselves, LinkedIn thinkpiece fodder writers, and academic career advice columns all fawn over this analogy. It shows up in the Science / AAAS career section, where a University of Kentucky physiology professor is quoted saying that running an academic lab is "equivalent to … running a small business out of the university." Lab Manager publishes pieces titled "Setting Up Your New Lab Like a CEO." Peer-reviewed lab-leadership literature has it too: PLOS Computational Biology, citing "the parallels between running a lab and directing a small company," goes so far as to suggest "the former may be even more complex."

The comparison flatters the PI as entrepreneurial and scrappy, and it flatters the entrepreneur-adjacent reader by treating research as "real work." I run an actual small business and have spent enough time around academic labs to find the analogy superficially appealing but structurally misleading.

"Essentially PIs are like captains of ships or CEOs of small companies. You've got to multi-task, you've got to get the most out of people and there is a lot of strategic decision-making involved."

— PI informant, in O'Kane et al., Technological Forecasting & Social Change (2020)

That quote is from a peer-reviewed study of 41 health-science PIs. It is the cleanest version of the analogy I have found, and it is wrong in ways the rest of this post will lay out.

Where the Analogy Feels True

The analogy feels superficially true because both roles require hiring, managing, and developing talented people. Persuasive writing is required to get money, whether the vehicle is a grant, a pitch deck, or any other financial proposal. In both cases, you are selling a vision to people who don't yet believe in it. Disciplined budgeting and project management keep the lights on in both scenarios as well.

Where It Breaks Down

The analogy breaks down when you start to look at second-order business problems.

Who Is the Customer?

In a business relationship, the customer pays you voluntarily and can leave. In academia, that customer doesn't really exist in a meaningful product-market sense. You could argue that trainees, peer reviewers, and program officers all consume writing products, but they are simply an audience to be persuaded. There is no real product-market feedback loop, retractions are rare for "oops, this experimental design was wrong," and cruft persists in the academic journal system.

The clearest demonstration of this gap comes from a PI who reaches for the business analogy and then defines it into something else:

"Establishing a lab is analogous to setting up a business. In a way, I visualize it as being given a small shop to rent in a big mall. In order for us to be noticed, we need to create a product (our data) that people (our colleagues) can look at and decide whether it is worth their attention/investment or not. We need to make a brand (our unique angle for producing data), find investors (therefore excite future potential reviewers/funding agencies), and gain visibility by going around and creating publicity (giving talks)."

— Valentina Greco, Mol Biol Cell (2014)

Look at the substitutions she lists: product = data, customers = colleagues, investors = reviewers and funding agencies, publicity = talks. None of those are markets. Colleagues pay attention, not money; reviewers gatekeep, they don't invest. The metaphor names the lab's outward-facing labor, but the moment each business term is translated into its academic referent, the property that made the original term meaningful (voluntary monetary exchange) is gone.

Who Bears the Risk?

When a PI fails to attain funding in a grant cycle, they generally keep their salary, their office, their title, and their institutional affiliation. A small business owner with a bad fiscal quarter has no such fallback: they make payroll from personal reserves, take on business debt obligations, and carry all liability when the lights go off.

The Science / AAAS Business Sense piece on starting an academic lab puts the framing crisply, quoting UPenn chemistry professor Jeffrey Bode:

"estimate your monthly expenses — your 'burn rate.'"

The burn-rate metaphor breaks the moment you ask whose money is running out. A startup's burn rate is the founder's and investors' capital being consumed. A lab's burn rate is drawn from a composite of grant funds and startup packages, and when those run out, the result is usually a fallow period during which the PI sticks around.

This is not only a critic's framing. Some of the same authors who reach for the small-business analogy will, in the same piece, concede that the salary structure is the opposite of small-business risk exposure. A Molecular Biology of the Cell essay that opens by calling the PI role "creative space to run a small business based on your own ideas" later writes:

"the relative personal financial stability of a research-intensive academic job: it is pretty unrivaled. … faculty members at our institution earn a significant part of their salary for their undergraduate teaching duties. … even those faculty members benefit from contracts that provide stability in the short term."

— Gladfelter & Peifer, Mol Biol Cell (2017)

Both halves of that argument can't be right. The thing that makes the professorial career attractive (salary continuity, contracts, research-direction autonomy, teaching-line stability) is precisely what makes it not a small business.

That said, salary safety is not a universal rule, and I should be honest about who it does not cover. Soft-money PIs at US medical schools, research institutes, health science institutions, and many non-tenure-track positions do have salary on the line:

"Grant writing is a huge pressure because I'm research only, my wife is research only and we have a mortgage so when you have to look after your own salary it can be tough."

— PI informant, O'Kane et al. (2020)

Soft-money PIs still don't sign personal guarantees on equipment leases, carry liability insurance, or face involuntary closure of the operation. But the gap between PI risk and small-business-owner risk is narrowest at this end of the spectrum, and "your salary is fine" is too strong a claim to make universally.

There is also a real overlap worth conceding: even tenured PIs with protected salaries describe a fiduciary obligation to keep grants flowing so that postdocs, technicians, and grad students don't lose their positions:

"We're always under pressure for money. … We've been living this way for 15 years. You have to constantly get on the treadmill, especially for the sake of the people in your group who are reliant on the money for a career."

— PI informant, O'Kane et al. (2020)

That payroll-driven pressure is a real point of contact with small-business ownership. The disanalogy is in the consequence, not the experience: when the funding does run out, the PI's house isn't on the line, but the postdoc's visa and the technician's mortgage are.

Revenue vs. Grants

Revenue and grants are not analogous. Grants are lump-sum, generally multi-year, and awarded on prospective work. Revenue is continuous, contingent on delivering something someone wants today, and tells you almost immediately if you're wrong. Calling grants "revenue" is the category error the whole analogy rests on.

Ken Gall, who has co-founded ten startups out of his Duke lab, is one of the few academics who have actually pitched both:

"It's hard to sell basic science to investors. … And it is very different from selling your work to a research grant committee. With a grant, you're laying out this meticulous plan that's reviewed behind the scenes and then they just tell you yes or no in a review document. With an investor, you talk face-to-face, and they give you a very small window to teach them why they should put their money in. And then they want their money back plus some."

— Ken Gall, in Duke Pratt's Hard-Earned Lessons of Academic Entrepreneurs

"And then they want their money back plus some" is the part the analogy keeps omitting. Grants don't expect a return; investments do. The accounting, the legal structure, and the relationship are different shapes.

Time Horizons and Failure Modes

Academic failure is slow, socially driven, and largely reputational. Business failure is fast, financial, legal, and somewhat reputational. Both are painful; they are not the same shape of pain.

Hiring and Firing

Grad students and postdoc trainees generally have institutional protections, degree requirements, visa considerations, and a fiduciary relationship with their PI and the PI's institution. In most US states, employees are at-will, paid market rates, and the relationship is bounded by job duties and performance. I concede that this is not the case globally, as labor protections outside of the United States can be significantly more robust. These are not interchangeable management problems.

"Scientific PIs do not normally have formal management training, and, instead, are typically self-made managers."

— PLOS Computational Biology, Learning How to Run a Lab

Being an untrained manager of trainees is not the same job as being an untrained manager of employees, either legally or in fiduciary terms.

"The main difference is that companies can use financial incentives, while scientific labs must rely mostly on career advancement and quality of the research experience."

— PLOS Computational Biology

That is the quiet admission that the compensation levers available to a business owner do not exist for a PI, who is largely bound by the rules of their grants and their home institution.

The most rigorous version of this argument I can find comes from a working PI in the O'Kane et al. study, describing what it actually feels like to run a lab without the management infrastructure a real small business would have:

"I find it quite ironic that I've got a million dollar grant with absolutely no training for how to run it. Sometimes I feel completely out of my depth or I don't know what I'm doing with the financial management. If I was running a small business then I would have people helping me to run it but as a PI I don't feel I have any training or support from the university at all."

— PI informant, O'Kane et al. (2020)

Read that quote twice. A working PI is saying, in print, that running an actual small business would come with more operational support than running a lab does. The standard analogy gets that backwards.

Accountability

Who can actually fire a tenured PI for poor performance? What even is the definition of poor performance in that context? And who, in turn, can fire a small business owner? The authority to hire, fire, promote, and pay differentially is what makes someone accountable as a principal of a business. Strip those levers and "CEO of your lab" is a costume, not a role.

Why the Analogy Persists Anyway

I would speculate that the analogy of the PI as a business owner persists because it gives the individual a sense of agency in a deeply bureaucratic system. It also gives administrators a little leverage to justify pushing operational burden onto faculty. For policy makers, the analogy offers a tidy frame for "innovation" and, more recently, for judging academic bodies of work on their societal and entrepreneurial impacts. Business audiences are met halfway by the invocation of familiar terms. The analogy is doing real social and rhetorical work even when it is structurally wrong.

There is also a vendor-marketing channel worth flagging. The Lab Manager piece I cited in the introduction, "Setting Up Your New Lab Like a CEO," is co-bylined with Cytiva (a GE Life Sciences brand) and closes, after several pages of "just like a business" framing, with the line: "GE can help with setting up the protein research workflow in your new laboratory." The lab-as-business framing makes a new PI legible as a small-business owner, and small-business owners buy capital equipment. The analogy isn't only a self-image and an administrative convenience; it is also, in places, a sales pitch.

It is not only outsiders making the comparison. Working PIs reach for it about themselves, and they reach for it most when funding is scarce. Bradley Olson, a PI and associate professor at Kansas State, in a BioSpace interview:

"On the academic side of being a PI, it is more like being a CEO at a small startup with the added responsibility of mentorship (in addition to university responsibilities)."

In the same interview, Olson notes that "funding for investigator-initiated research has been going down over the last 10 years or so. It is really making competition fierce." Those two observations belong together. The "scrappy entrepreneur" frame is more attractive when grants are harder to win, and it lets a PI describe their increasingly bureaucratic, increasingly precarious workload as something heroic.

The cleanest test of the analogy is what happens when academics actually try to be small-business owners. The Duke Pratt School of Engineering profiled four faculty across what their tech-transfer office calls a "continuum of involvement." Ashutosh Chilkoti has spun five companies out of his lab and prefers what he calls the "light touch" model. Note who runs the resulting companies:

"Chilkoti advises the companies but is not part of their C-suite level teams. … Technology applications are usually championed in the lab by grad students and postdocs, who then become the leaders of the spinout, like PhD graduates Kelli Luginbuhl steering Isolere and Stefan Roberts helming inSoma."

— Duke Pratt, The Hard-Earned Lessons of Academic Entrepreneurs

When a PI's technology actually spins out into a company, the operating role usually goes to a former trainee, not the PI. The faculty member who did take the CEO seat at her own startup, Sonia Grego, describes the experience this way: "We really did not sleep much. … We're building the plane as we fly it. … nobody cares about the company as much as I do." That is the unmistakable register of an actual small-business owner. And it is exactly what most academic founders, given the choice, opt out of.

There is one last way the analogy survives: by quietly downgrading itself. Derek O'Keeffe, in a recent Science / AAAS Working Life essay, describes how an MBA reshaped his lab. He is enthusiastic about applying business-school methods, and along the way he writes:

"Not everything translated perfectly from business school to academia, of course. Unlike a business, we're not focused on making a profit."

"We began to enlist our 'customers' — doctors and patients — early on, testing our technology with them and adjusting it based on their responses."

He concedes the load-bearing disanalogy in one sentence (no profit motive) and reaches for scare quotes the moment he uses the word customer literally. He is using business-school technique (communication norms, mentorship structure, user research) and calling it business-school identity. Those are different claims. The techniques generalize; the identity doesn't. This is exactly how the "PI as small-business owner" framing survives contact with reality: the operator quietly downgrades it to "PI as person who learned useful things from business school," which is not the same claim and is not contested by this post.

Where the Analogy Actually Fits: Core Facilities

The small-business analogy isn't wrong about academia. It is just misapplied. The role inside a university that genuinely resembles a small business is running a core facility: a shared sequencing, imaging, cytometry, mass-spec, or HPC center, not a PI's lab. Core labs have real customers, both internal and external to their host institution. Outside of financial or reputational incentives, there is no obligation for any group to keep working with a particular core, and PIs are free to walk away from a core-lab relationship at any time. Core labs have to manage real revenue and sit down regularly with institutional accountants to defend a budget. A PI's multi-year grant can buy time and resources at a core, and reputation can win a core large external partnerships. This is where service and expertise actually behave like a product in a concrete, paid-for sense.

Recharge rates are reviewed, and at federally funded institutions they are governed by Uniform Guidance / OMB principles. That is closer to the regulated-utility end of "small business," but the operational logic (set a price, deliver a service, justify your costs) is the shape of a business, not a research program. Most cores are required to roughly break even across a rolling window. Persistent under-recovery or over-recovery is a problem you have to actively manage, the way a small business manages margin.

Core directors typically supervise staff scientists and technicians as employees, not trainees. The legal and fiduciary shape of that relationship is the employer/employee one this post argued PIs don't have with grad students and postdocs. Compensation for these employees lags behind private-sector rates and may not retain talent the way a true business could, but that is a different problem entirely. The product-market signal is still real: if the queue empties, if commercial vendors undercut rates, if users stop rebooking, the feedback arrives on a much faster timescale than any R01 cycle.

The point isn't to flatter core directors; it's to sharpen the original claim. When people reach for "running a lab is like running a small business," the example they are half-remembering is usually a core facility, or a contract research org embedded in a university, not a PI's research group. Conflating the two lets the rhetoric of "you're basically a CEO" attach to the role with the least actual customer exposure.

What a Better Frame Might Look Like

An academic PI may be closer to a department head at a large nonprofit: someone with budget authority but no profit-and-loss responsibility, working under a mission-driven mandate, accountable through committee governance. Or perhaps the role is closer to a skilled-trade guild or union: apprenticeship-based, reputation-driven, producing output that is credentialed people as much as it is artifacts. Neither framing is perfect. The point is simply that "small business" isn't a uniquely apt way of describing the role of a PI and its responsibilities.

A sharper analogy comes from the same PLOS Computational Biology piece that also reaches for "small company":

"a small software company very much resembles a scientific lab: It is a small motivated group of highly-educated professionals creating nonmaterial products."

— PLOS Computational Biology

This is a sharper comparison than "small business" in general, but notice what it smuggles in: the shared property is team shape, not market exposure, legal liability, or compensation structure. Team-shape similarity is the thing the analogy consistently gets right, and the thing people keep mistaking for a full isomorphism.

What This Means in Practice

  • PIs leaving academia: your skills overlap less than you think. Plan accordingly.
  • Entrepreneurs founding academic-adjacent spinouts: the incentives you are used to don't apply on the other side of the wall.
  • Administrators: stop using "you're basically running a small business" as a motivational framework when you have not ceded the authority or risk that would make the statement genuinely true.
  • Funders: if you want PIs to operate like businesses, give them some form of customer-like feedback loop, not just review panels and budget spreadsheets.

Parting Thoughts

A note on where I am writing from. I have never been an academic PI, though several of my close friends are, and I have heard a lot of these conversations from the inside-adjacent seat. I have worked in core labs (the Arizona Genomics Institute and CyVerse), which is part of why the "Where the Analogy Actually Fits" framing rings true to me: those operations behaved more like small businesses than any PI's lab I have spent time in. Today I run an actual small business, which is the other half of why I find the standard analogy frustrating. None of that makes me right, but it is the vantage point this post is written from.

I respect both crafts. This is a critique of the comparison, not of either group. Running an academic lab well is difficult. Running a business well is difficult. They are difficult in different ways, and conflating the two roles cheapens both.

Source Material

Pieces that explicitly make the PI-as-small-business-owner comparison, ordered roughly from most quotable popular-press target to most useful academic framing:

Have thoughts on this analogy? Find me on Bluesky or LinkedIn.

Note: This post reflects my own experiences and perspectives. In all transparency, I used Claude (Anthropic's AI assistant) to help with link population, grammar checking, and spelling corrections, not for content generation or overall structure.

About the Author

Dr. Ryan Bartelme is the founder and principal data scientist at Informatic Edge, LLC. His expertise spans bioinformatics, data science, microbial ecology, and controlled environment agriculture.