In this episode of Enterprise Sessions from the University of Bristol, Professor Michele Barbour speaks with Professor Paul Bates, world‑leading expert in flood inundation modelling and co‑founder of Fathom, one of the University’s most successful research‑driven companies.
Paul reflects on a remarkable career that began with a Bristol PhD in the late 1980s and evolved into pioneering work that transformed global flood modelling. He describes the technological shift that enabled a new generation of high‑resolution terrain data, the academic debates that reshaped the field, and the multidisciplinary collaborations that built the foundation for Fathom’s modelling techniques.
The conversation traces Fathom’s origins from two ambitious PhD students with an idea, through early years of bootstrapping, to international clients including insurers, banks, multinationals, and the World Bank. Paul also discusses the challenges of spinning out before universities had mature commercialisation systems, the importance of staying ahead of competitors through transparency and innovation, and the recent acquisition of Fathom by Swiss Re.
Finally, Paul reflects on what research entrepreneurship means within academia, how Fathom has strengthened Bristol’s scientific capabilities, and what lies ahead for both him and the next generation of global flood models.
🔍 In the episode:
· The origins of flood inundation modelling at Bristol
· How new airborne laser mapping transformed what was scientifically possible
· Overturning long‑held assumptions in the field
· The multidisciplinary team behind high‑resolution flood models
· Serendipity, road trips — and how two PhD students sparked a company
· Fathom’s unconventional path: bootstrapping, grants and early customers
· Data‑as‑a‑service before it was mainstream
· Building global flood maps used by insurers, governments and financial institutions
· Staying ahead of competitors by publishing methods openly
· Growing from four founders to a 50‑person global team
· Acquisition by Swiss Re and what it means for the future
· Entrepreneurship in academia: culture, opportunity and barriers
· The virtuous cycle between research and commercial innovation
· What’s next: NASA’s SWOT satellite and the next era of global flood modelling
🌐 About the Enterprise Sessions
The Enterprise Sessions bring together a diverse mix of company founders and researchers who talk openly about their personal experiences of forming spinouts and start-ups, raising capital, academic-industry partnerships and the joys of translating research discoveries into real-world impact. The series aims to inform, inspire and challenge myths and stereotypes about research commercialisation and how businesses and universities can work together to tackle society’s biggest challenges.
👍 Like, Share, Subscribe, Explore
If you found this episode inspiring or informative, please don’t forget to like and share. Visit our website or subscribe to the University of Bristol’s YouTube channel for more Enterprise Sessions.
https://www.bristol.ac.uk/enterprise-sessions
Paul Bates – LinkedIn
Michele Barbour – LinkedIn
Chapters:
- 00:00 – Introductions
- 00:53 – Arriving in Bristol and the early days of flood modelling
- 02:11 – When new terrain data changed everything
- 04:12 – Interdisciplinarity before it was fashionable
- 06:34 – The epiphany: better data beats better physics
- 07:57 – Challenging a half‑century of scientific assumptions
- 09:37 – Industry collaborations and glimpsing commercial potential
- 11:00 – A road trip, two PhD students, and the birth of Fathom
- 13:14 – Trying—and failing—to spin out
- 15:32 – Bootstrapping with grants, market research and early clients
- 17:56 – Building early products and standing on their own feet
- 19:43 – Paul’s role: strategic guidance and academic expertise
- 20:55 – The virtuous circle between research and commercial products
- 22:14 – What Fathom sells: global flood maps & risk models
- 24:55 – Who buys it: insurers, regulators, banks, tech giants, NGOs
- 26:49 – Why governments are a missed opportunity
- 28:18 – Competition, transparency and staying ahead
- 29:49 – Innovation, data, computing and continuously improving accuracy
- 31:00 – Why Fathom stayed in Bristol
- 32:12 – Investment, scaling and the climate services boom
- 33:25 – Acquisition by Swiss Re
- 34:47 – Entrepreneurship in academia: what’s changed and what hasn’t
- 38:04 – Role models and culture change
- 40:00 – Looking back: what younger‑Paul would think
- 42:14 – What’s next: NASA’s SWOT mission and the future of global flood models
Transcript:
00:00:09 Prof Michele Barbour
Welcome to another enterprise session from the University of Bristol. I’m Professor Michele Barbour and today I’m very happy to be speaking to Professor Paul Bates. Paul, I’m really grateful for you sparing the time to speak to me. Thank you very much.
00:00:19 Prof Paul Bates
That’s quite all right, Michele. Glad to be here.
00:00:21 Prof Michele Barbour
So I would love to explore the Fathom story. Quite often when I speak to academic entrepreneurs, they’re just in the process of setting up a company or maybe they’re planning to do one next year. Yours is a mature and very successful company, so it’ll be quite a different flavour, I think, looking at how things were for you and how they’ve gone and what’s changed since.
But where I usually like to start is before that. Can you tell me a little bit about your background, maybe what you were doing before you came to Bristol, and what brought you to Bristol in the first place?
00:00:53 Prof Paul Bates
So I came to Bristol in 1989 to do a PhD in flood inundation modelling, and basically I’ve stayed here doing that ever since. I am a one-trick pony, but it’s a good trick.
00:01:05 Prof Michele Barbour
It is a good trick. You might be surprised, but a lot of people come to Bristol and find it quite a sticky place — whether it’s the city, the university, or their research.
00:01:15 Prof Paul Bates
I started doing a PhD in flood inundation modelling, then did a postdoc here. I was funded by the US Army Corps of Engineers for my first postdoc, but I was working in Paris for Électricité de France at the Laboratoire National de Hydraulique. From there I did another postdoc and eventually came to a faculty position here at Bristol.
At that particular point, I’d pretty much exhausted all the inundation modelling — and the science — that we could do, because the data sets just weren’t available to take the science any further. So I put it aside for a few years, until the end of the 1990s and early 2000s, when new data sets coming on stream suddenly made us realise what we’d be able to do. We started building new models with these incredibly rich data sets — mostly terrain laser mapping from aircraft.
00:02:08 Prof Michele Barbour
How did it come about that there wasn’t enough data and then the means to collect it became available? Was it that the motivation became more prevalent?
00:02:16 Prof Paul Bates
The means to collect it — the technology moved on and allowed academics to flood in behind and grab that data and do new science with it. We wanted in this country to map our flood risk better, and it was realised that one of the limitations to doing that over large areas was our methods of collecting terrain data. We used to send people out in the field with a survey level, which is slow and cumbersome. We had some ways of doing it from aircraft using techniques we’d had since World War II, but again they’re slow, fiddly, and require a lot of manual work.
Then military technology for mapping the ground from above became available, which involved laser scanning from aircraft. That’s much more automated, with less human intervention, and it collects a lot of really accurate data very quickly over large areas.
00:03:08 Prof Michele Barbour
That’s a gold mine for an application.
00:03:09 Prof Paul Bates
So we suddenly went from a situation where, if you look on a survey map, the rivers and floodplains appear quite obvious because they don’t have any contours — we just didn’t know what the terrain looked like. In the past we thought they were flat, but in reality they have this complex surface micro-topography that you need to pick up to model flood risk. So we went from this almost terra incognita of traditional mapping to having incredibly rich data sets with a data point every metre or two to a few centimetres of accuracy, over tens of kilometres of river. And this stuff appeared almost overnight.
00:03:48 Prof Michele Barbour
Amazing. And — forgive my ignorance — this is not the surface of the river, which is going to be more or less flat because water is self-levelling. You’re looking at the land adjacent to the river that water would get into if levels rose?
00:04:05 Prof Paul Bates
Exactly.
00:04:06 Prof Michele Barbour
So you had the initial interest, you had to park it a little bit because the data wasn’t available in sufficient quantity and quality, and then the data became available in the late nineties — all your Christmases at once, presumably. What happened next? And given this was a rapidly developing situation, how did you get it funded?
00:04:27 Prof Paul Bates
We had some research grants ongoing and we diverted some of the resource toward exploring these new data sets — we hadn’t necessarily declared that in the grant applications. Proof of concept was developed, which allowed us to acquire other research grants and hire PhD students to do some of this work. We were able to resource it pretty quickly and made rapid progress that landed well within the literature. So we were up and running very quickly.
00:05:04 Prof Michele Barbour
What’s your angle on this — are you a geographer, an engineer, a computer scientist? I can see how all of them are relevant. And who was ‘we’? What other skills did you need around you?
00:05:17 Prof Paul Bates
I am a geographer. And geographers are often jacks of all trades — we’re not mathematicians, but we’re okay with the maths; we’re not computer programmers, but we write good computer code; and we’re not full-blown social scientists, but we do some really great social science.
The other people in the research group were physicists, mathematicians, engineers and geographers, so bringing that set of people together around the common endeavour was how we got it moving.
00:05:53 Prof Michele Barbour
And this was at a time when interdisciplinary research was not anywhere near as common or fashionable as it is now. Was it unusual to have a physicist and a geographer and a mathematician all sitting around the same table talking about a project?
00:06:10 Prof Paul Bates
It didn’t seem so. Other parts of geography had similar things going on — I think geography was fairly open to having people come and work with us. We’re not very disciplinarian and we’re used to working across disciplinary boundaries, so it seemed very natural and normal.
00:06:31 Prof Michele Barbour
So tell me what you built. You’ve got these huge data sets. What did you make with them and what were you ultimately trying to achieve?
00:06:42 Prof Paul Bates
With the new data sets, we went back to all the questions we’d had during my PhD that we were never able to answer — around whether we could build computer models that predict the areas at risk of flooding. We’d exhausted our ability to build and test those models by about 1993-94, and the new data allowed us to start that process again.
We also realised we needed a new class of models, because the data was so dense and rich that the models needed to be much more efficient and simpler in order to ingest it all. We tried to work out what matters more for improving a model’s skill — is it more physics, or is it better accuracy and resolution of input data? And it turned out it was the latter. If you don’t have good data, you don’t get good results. That rather overturned the paradigm of flood modelling for the previous half-century. For a long time — myself included — we thought we’d make our models better by improving the physics. That turned out not to be the case. You make them better by making the grids finer and the terrain data more accurate.
00:07:53 Prof Michele Barbour
So that was quite an epiphany for you. But presumably you were still surrounded by many in your discipline going with the dogma of the last fifty years. Did the whole field flip almost overnight, or did you have to fight for it?
00:08:08 Prof Paul Bates
I think that paradigm shift has now been firmly accepted. At the time, we had to deal repeatedly with the same kinds of review comments, constructing arguments to bat them back and persuade editors to publish our work. Over time we managed to persuade our colleagues that things could be done differently, and the view that developed in and around Bristol of how to do flood modelling is now much more widely accepted.
00:08:47 Prof Michele Barbour
Some people might relish that challenge — overturning something, showing people how it really works. Others would be put off by having to answer the same questions over and over with journals rejecting their papers. How did you find it as a researcher, and at that point quite an early-career one?
00:09:11 Prof Paul Bates
I quite enjoyed it, actually. I enjoyed the challenge of arguing back and winning the battles. And still do.
00:09:23 Prof Michele Barbour
Great research with big potential impact — and before impact was such a strong part of the narrative as it now is. This could quite happily have stayed as research. You didn’t need to take it into a commercial endeavour. So tell me a little bit about the why — and then we can explore how you went about it.
00:10:10 Prof Paul Bates
Let’s do it the other way around — let’s explore the how and that will make the why clearer. The how is that if you’ve developed an ability to model risk with reasonable skill over large scales and in granular detail, you have a set of techniques that the insurance industry starts to be interested in. So increasingly during the 2000s we started to connect with insurance industry collaborators, and they opened our eyes somewhat to the commercial potential of this work. After being part of something called the Willis Research Network and having PhD students funded by the insurance industry, we were much more aware of where our techniques could fit in and what their value might be.
And the why is that I had a couple of PhD students — one funded by the insurance industry, called Chris Sampson, and another funded by the AHRC, called Andy Smith. They were on a road trip around Las Vegas and Nevada prior to joining me at a conference in San Francisco. Whilst they were on this road trip, they had an idea — as they were coming to the end of their PhDs — that they weren’t going to pursue a standard postdoc, and that the things they were developing had commercial value in the insurance sector.
00:11:44 Prof Michele Barbour
The value of a road trip! What was your gut reaction when they told you?
00:11:51 Prof Paul Bates
I was infected by their enthusiasm. I said, ‘Great, let’s go for it.’ And then we brought in my colleague Professor Jeff Neal from geography, and the four of us founded the business.
00:12:05 Prof Michele Barbour
I love it. And what’s particularly interesting is that it was your PhD students who came to you and said let’s form a company — almost the opposite of the usual story where an academic with an entrepreneurial leaning influences their early-career researchers.
00:12:33 Prof Paul Bates
Yes — they had the time, the risk-taking appetite, and the sufficiently devil-may-care attitude to think, ‘We’ll start a business, probably it’ll fail, but we’ll learn a bunch of stuff on the way and then we’ll get a corporate job.’ But perhaps to make the business successful, they ought to bring in some older heads to help a little. Hence the four of us formed the team.
00:12:58 Prof Michele Barbour
So the four of you formed Fathom. In 2013 you formed the company. Nowadays forming companies out of university research has become much more common, but 2013 was a very different time — and the ecosystem inside and outside the university was at a different stage of development. How did you find that process? How was your proposal to form a company received?
00:13:38 Prof Paul Bates
The first thing I should say is that you have to gain permission from your head of school and ultimately your dean in order to spend some time forming a company. I was very grateful to Tim Gallagher, who kindly looked at the proposal and said, ‘Yes, this looks interesting — you can spend maybe half a day to a day a month doing this.’
But the rest of the university was deeply frustrating to interact with. The commercial team seemed to have one model of spin-out, which was that a widget had been developed, the university owned all the IP, and therefore could call all the shots. We were developing a data-as-a-service business, which didn’t fit that model and I’m not sure was quite understood by the commercial team at the time. So they tried to impose the widget conditions on us.
We wanted to be a spin-out, for the university to take a stake and for us to have the benefits that would bring. But after a couple of months of negotiations that went nowhere — in fact went backwards in many ways — we gave the university an ultimatum: the IP is already in the public domain, you don’t own it, our competitors are using it. Either we come to a sensible agreement, or we’ll be a start-up instead. They didn’t respond, and we started up rather than spinning out.
00:15:19 Prof Michele Barbour
Universities at an earlier stage in their commercialisation journey can have quite a narrow definition of what a spin-out can be. Things like the USIT guidance are so valuable because not every university has to figure it all out for themselves. I’m happy to report that things have changed a great deal.
00:16:10 Prof Paul Bates
I have been told.
00:16:11 Prof Michele Barbour
Good — if you ever want to do another one! But despite those challenges, in 2013 you formed your company. Given this isn’t a widget and doesn’t require a wet lab, what did you actually need? Did you need investment, funding?
00:16:35 Prof Paul Bates
We did need some money to cover computing and to provide at least basic living expenses for Andy and Chris so they could do the bulk of the work. They took the bulk of the shares as well — they took the risk, they put the effort in, so the two academics were the minor partners in the arrangement.
We were lucky enough that at the time, the Natural Environment Research Council were giving out impact awards. Through the university’s NERC Impact Accelerator Account, we got £25,000, with which we did some market research and proof of concept — it paid for Andy and Chris to go to London and talk to insurers, to make sure that what we thought was a good idea was something people might actually pay money for.
Then we applied for a further grant from NERC and eventually received around £200,000 from their strategic impact funding round. That paid for postdoctoral salaries for Andy and Chris, plus some travel and more market testing and prototyping. That took us through the first eighteen months. By the end of that, we had the first products and the first commercial contracts, and we had enough money to rent a tiny little office down at Engine Shed and pay basic salaries for Andy and Chris to get going.
00:18:39 Prof Michele Barbour
So you didn’t have to go down the route of venture capital — you bootstrapped it, to use the modern term.
00:18:48 Prof Paul Bates
We bootstrapped for the first seven years of our existence, which was a really, really good idea. We got Research Council grants that got us moving, and we also received a gift from Google — at the time they were giving gifts of up to $100,000 US to interesting projects they wanted to promote. So with Research Council money and a gift from Google, we had enough seed-corn funding to get going. By the time that came to an end, we had contracts in place.
00:19:24 Prof Michele Barbour
You could stand on your own two feet. And your PhD students Andy and Chris were really building and driving the company forward, while you and Jeff were still involved. What role did you take?
00:19:45 Prof Paul Bates
All the way through, I’ve chaired the company board. In the early days that was just chairing a monthly meeting — providing strategic advice on what we should do, how we should develop the software, what might be of commercial interest. So the academics didn’t do any of the work, but we provided advice and contacts, and were occasionally wheeled out when meeting people who might want to buy the data, as a signal that ‘here are the experts from academia, this is where we come from.’ Not too much of an input, but hopefully an effective one.
00:20:25 Prof Michele Barbour
Helping drive strategy and being a sounding board for two people who were clearly very skilled but less experienced. Has that long-standing relationship with the company had any influence on how you approach your research?
00:20:55 Prof Paul Bates
Absolutely. The company activity is completely symbiotic with what we do on the research side — mutually reinforcing. We do some fundamental research in the university, put it in the public domain, some of it gets picked up by the company who turn it into global products, and we can then pull back and license those back into the university to fuel the next phase of research. It’s a really virtuous circuit.
We’ve been able to write papers over the last seven or eight years that we would have had absolutely no possibility of writing from a university research group funded by standard grants. If I went to the Research Councils and said I want to build a global flood model requiring thirty postdocs for ten years and a million pounds of computing, they would just laugh. But that’s what we’ve used to build the products we’ve got. What we’re doing is leveraging commercial money to build the science team we always wanted and could never have achieved any other way.
00:22:08 Prof Michele Barbour
Before we go on to what that science team is going to do next, tell me a bit more about the product. I’m getting that this is about flood modelling, and you’ve mentioned the word global. What exactly are you selling, and to whom?
00:22:24 Prof Paul Bates
There are three main products. First are flood maps — global maps at thirty-metre resolution. So every thirty metres over the entire terrestrial land surface, we can tell you, for a given probability flood event, what the water depth would be at that location. Think of it as street-level mapping. That allows you to say: if I have a set of assets, which of those are at risk and by how much?
You can also model different scenarios — we work on the probabilities of events. So if you were to have a flood event in Bristol that was statistically likely to occur only in one per cent of all years — once every hundred years on average — we can tell you what the water depths from that event would be.
00:23:52 Prof Michele Barbour
And who’s using that information? Who are your paying customers, generally speaking?
00:23:55 Prof Paul Bates
Surprisingly, the market for flood hazard and risk information is much broader than I ever thought it would be. These hazard maps and associated risk models are used intensively by the insurance industry for pricing — both insurance and reinsurance — and by insurance industry regulators, because insurers have to prove they’re capable of being solvent if really bad things happen; they have to hold sufficient capital to pay out if a series of unfortunate events were to occur simultaneously.
But that kind of information is now increasingly being sought by financial houses, banks, and other financial institutions. If you have a global book of mortgage contracts, you might want to understand what the risk to those is. If you’re investing in long-term infrastructure assets, you might want to know the risk to those now and into the future. So as well as the major insurance players, we have clients among the big four public financial services firms, large finance houses like BlackRock, multinationals like Microsoft, and NGOs as well. Our data are licensed by the World Bank and used to inform investment decisions in particular regions and the risks to those from flooding due to climate change.
Selling to governments is much harder — they’re quite opaque and monolithic organisations and it’s difficult to know who would take the decision to purchase your data. So although it’s an ambition, we haven’t had a lot of success there yet.
00:26:26 Prof Michele Barbour
You’ve got willing customers who are paying, while chasing the ones that would be ethically productive but commercially difficult.
00:26:26 Prof Paul Bates
Sovereign states could be really valuable. A developing country might not have the resources — the university staff or the engineering staff — to develop their own flood hazard maps and put them on a website for their population. If they went to a consulting engineering firm, they’d be charged a fortune for a bespoke product. Yet they could license the Fathom data for a reasonable sum — a fraction of what a consultancy would charge — and within six months have a flood map that you could type your postcode into and see the flood risk around your house. That would go a huge way to increasing awareness of risk, allowing it to factor into planning decisions and risk preparedness.
I’d like it to be Fathom’s data, but as long as it’s good data, I don’t mind. We have a series of competitors whose work I respect — a small number who’ve been in the game a long time and produce good stuff — and there are some charlatans I have less time for. But the data could do an awful lot of good getting into the hands of governments and the public domain.
00:28:18 Prof Michele Barbour
How have you dealt with the risk of competition? Other people have access to the same data, and there are other smart geographers and programmers out there. What’s to stop someone doing what you do and selling it for slightly less?
00:28:38 Prof Paul Bates
Fathom publishes its methods openly in peer-reviewed journals — so literally, anyone can read what we do and if they’d like to copy it, they can go right ahead. It’s quite difficult to build flood models, though. You need a certain degree of technical expertise and a large team. You can read our methods and you’re welcome to try to replicate them, but it’s hard. It’ll take quite a lot of investment over a period of time and you might still not get a good result. And by the time you do, we’ll be another eighteen months or three years on and we’ll have done the next thing already. So we’re just backing ourselves to stay ahead.
00:29:25 Prof Michele Barbour
And it’s worked! What have you had to do to keep modernising and innovating over the years — greater resolution, extending to new geographies?
00:29:49 Prof Paul Bates
Two obvious things drive it forward. First, better data sets come on stream all the time through new satellites, new data availability, and different ways of processing them — principally in the last few years using machine learning, which has improved the input data. Computing power has also advanced, allowing us to deploy more power and build finer resolution models.
Second, we’ve been able to innovate in terms of numerical techniques and refine how we go about the modelling itself. Continuing the academic research and pulling that through into the business side — those things together allow us to innovate continuously. For the customer, this means the models are more accurate and more granular over time.
00:30:54 Prof Michele Barbour
Over the roughly twelve years the company’s been in operation, you’ve stayed in Bristol. That makes me extremely happy. Why, and how have you managed to resist the temptation to move to London or the US?
00:31:15 Prof Paul Bates
London’s a main market, but it’s not far away, and we do have sales and marketing teams there — since we’re now part of Swiss Re, we have offices in the Gherkin, which is rather nice. But the technical team is all based here in Bristol, because there is a university here with a fantastic flood modelling group and a ready supply of talent, and the interaction between Fathom and the university team is very creative. And Bristol is a lovely place — it’s an easy place to attract people to and persuade them to stay.
00:32:00 Prof Michele Barbour
We talked about the early stages of Fathom and how you got it running. Bring us a little bit more up to the modern day, because I know you had some big news last year.
00:32:17 Prof Paul Bates
We bootstrapped the company until about 2020. At that point we realised that the market for climate services firms was really taking off, with a lot of new entrants — normally with rather sketchy techniques with the veneer of AI or machine learning, but essentially black boxes. Nevertheless, the marketing was great, and we realised we needed to accelerate our growth non-organically to stay ahead and demonstrate real ability.
So we took investment from Moody’s, the ratings agency, who took a stake in the business. We used that to grow the team from about twenty people to fifty over eighteen months. The new products we built and the new marketing verticals and sales teams we put together allowed us to grow the financial side of the business over the following years.
With that, we attracted the attention of Swiss Re — a four-billion-US-dollar-a-year-turnover Swiss reinsurance behemoth. In 2023 they approached us about acquiring the business completely, the timing was good, and we had a series of negotiations that culminated in an exit in December 2023.
00:33:51 Prof Michele Barbour
Congratulations — that’s the success story so many founders picture and so many don’t achieve. Are you still involved?
00:34:02 Prof Paul Bates
I’m still chairman and still working two days a week for Fathom. Everybody who was involved at exit is still involved in the company. Nothing’s really changed apart from who owns the shares.
00:34:13 Prof Michele Barbour
Your start-up was formed in 2013, mine in 2014 — we’ve established the world was in quite a different place back then. What about the more general culture at the intersection of academia and entrepreneurship — do you think that has changed since 2013? Is it more normal?
00:34:46 Prof Paul Bates
I don’t think I see a change in the rates of academics forming companies, or the propensity to do so. Maybe I’m looking at a small or biased sample, but I don’t get a sense from colleagues here or elsewhere that more people are thinking about doing this than previously. I don’t think in the past people really thought about it, and I don’t think they really think about it now. That’s a shame, because it’s probably an easier thing to do now than it was, and there is probably more support. But it’s not first and foremost in what most academics think of as their career arc.
00:35:32 Prof Michele Barbour
Numerically, back in our day the university typically put out one or two spin-outs a year. Now it’s between five and ten. So it has increased, but not uniformly across the organisation — some of our biggest growth recently has been in social enterprises. I’ve identified certain disciplines where there has been a really rapid rate of growth and others where there’s been little or none. Whether that’s because of the opportunities available, or more a cultural aspect of how different parts of the organisation see entrepreneurship, I’m not sure.
00:36:31 Prof Paul Bates
I think two things are going on. Firstly, in engineering and biotech, there is more investment, but equally there are probably more exemplars. If you’re a junior academic, you might look around your department and see a few people who’ve started up businesses and have role models you can follow. And those role models breed further interest — but there have to be a sufficient number of them, and they have to be visible.
As you said, we’ve gone from one or two start-ups a year to five to ten — an order of magnitude increase — but with four and a half thousand faculty, the change is so small as an ordinary academic you wouldn’t notice it. So empirically, I don’t see any change in the propensity of my colleagues to start businesses, and even given the order-of-magnitude increase you’ve described, I wouldn’t expect to, because it’s pretty rare anyway.
00:37:39 Prof Michele Barbour
It is rare, and one of the joys of my job is that I’m not trying to convert every academic into wanting to form a spin-out. What I would like is for a greater portion of our academic and research staff to see the point — whether they do it or not is secondary, but to see that it has a place in the academic endeavour and in delivering impact from research. Not to convert them, but to help them see the value. How can we tell those stories and help people understand?
00:38:31 Prof Paul Bates
Role models are important, and highlighting them more. I also think challenging people more to at least have the thought: if you were starting a business, what would it be? Maybe some workshops — spending some time going through that process, even if you’re not persuading everyone to follow through. The act of just thinking through what it might be could be quite instructive and might increase conversion rates.
00:39:01 Prof Michele Barbour
And one thing that struck me as you were speaking earlier is that your insight that your research had a commercial application arose through collaborating with industry — which is the other half of my job, encouraging researchers to actually collaborate with companies. In a sense, it sounds like that was a pipeline for you.
00:39:34 Prof Paul Bates
Yes — it took what was a vague idea to ‘I think we can see how this would actually work.’ We had enough insight into the business side to know where the gap was, and that we had techniques that could probably fill it.
00:39:51 Prof Michele Barbour
The way you tell your story, it’s very coherent — every single step seems logical. Did you have this in your mind when you started out? Is this something you pictured when you first came to Bristol all those years ago?
00:40:18 Prof Paul Bates
Not then. Whilst I was a junior faculty member, I saw my PhD supervisor start a business, which was quite instructive — I saw it was quite fun for him, which was good to watch. So I had at least the idea that this was something that could be done.
But I think a lot of the logic and coherence is retrospective. The real spark was the serendipity of having two really amazing PhD students at the right time, in the right environment, with the business opportunity there for the taking, and then them being prepared to take a risk.
The other thing that prevents academics following commercial opportunities is that it is a risk — to carve out the time to do it, you might have to reduce your time at the university and take a salary hit. If you’ve got kids and a mortgage, that’s not going to land well when you go home and explain the plan to your partner.
00:41:22 Prof Michele Barbour
That’s why we have enterprise fellowships — a partial solution, but I’d like vastly more resource to support exactly that. So given that you observed your PhD supervisor set up a company, if that earlier version of you could look through time and see where you are now and everything that your own PhD students have done, what would he say?
00:41:49 Prof Paul Bates
I think he’d be really happy. I only ever do stuff that I think is going to be fun, and I don’t plan much more than that. My two main aims are to have fun and to try and build the best flood models in the world. I don’t think too much further than that.
00:42:11 Prof Michele Barbour
So what next — what next for Fathom, and what next for you? Do you have another innovation? Do you want to do another start-up, or are you done with that?
00:42:22 Prof Paul Bates
I’m not sure I’ve got another start-up in me, but you never know — it’s certainly easier to do a second time, having seen the first. You make fewer mistakes.
On the academic side, I’m on the science team for a NASA satellite mission called Surface Water Ocean Topography, which launched in December 2022. I joined the initial team that refined the concept back in 2004, so I spent twenty years helping get that satellite launched and flown, and we’re now just starting to get data from it. I’ll use those data in my academic research, but we’ll also use them in Fathom to drive the next generation of global flood models with even better accuracy and more detail. So that’s my focus for the next couple of years.
00:43:14 Prof Michele Barbour
That’s incredibly exciting. Professor Paul Bates, I can only thank you. This has been a really fascinating interview — I feel like I’ve learned a great deal. I’m both fascinated by your journey and relieved that so much has changed since 2013. I don’t think we’d have as many spin-outs as we do if things were still as they were at that time.
00:43:38 Prof Paul Bates
That’s great. Thank you very much, Michele.
00:43:40 Prof Michele Barbour
Thank you, Paul.
Leave a Reply