Rob Carlson Rob Carlson, co-founder of Bioeconomy Capital and author of “Biology is Technology, speaking about the future of DNA synthesis at SynBioBeta 2018.
Home » Policy & public » Biosecurity » How to win in the bioeconomy: A peek at Rob Carlson’s playbook

How to win in the bioeconomy: A peek at Rob Carlson’s playbook

The bioeconomy is booming, and most predict it will grow even more in the coming years. In a field that changes and learns new things daily, it can be hard to wrap our minds around not only the future of synthetic biology, but also which technologies to invest in (both in terms of human and financial resources) and which are simply hype.

Rob Carlson is co-founder of investment firm Bioeconomy Capital and author of “Biology is Technology”. We sat down with Carlson to pick his brain about the latest revenue technology data from his Bioeconomy Dashboard and what trends are tickling his interest these days, covering topics from the synthetic biology stack to some striking industry trends.

In “The Future of DNA Synthesis” panel discussion at SynBioBeta in October, there was a question whether recent announcements from IDT, Genewiz, and Twist Bioscience signalled some change in the synthesis business. You didn’t seem convinced of this.

Looking at the DNA synthesis market over the last 20 years, how long companies tend to stay independent, and the size of the market, it’s just a really tough business to be in. While the demand for synthetic DNA, and synthetic genes in particular, is certainly growing, the price per base is falling so fast that the total value of the market is stagnant. New companies are trying to grab a share of a revenue pie that isn’t growing much, if at all. In that context, you have to have real technological or economic advantage just to attract customers, and even then it isn’t clear that any one company can dominate the supply chain.

I keep seeing reference to “the winner” in biological engineering and manufacturing, and I’m not sure we’re going to see a single winner. Across the bioeconomy, the barriers to entry seem to be coming down, and the markets are large. It isn’t obvious that there will be an Intel or a Microsoft of biology. More likely, we’ll have a larger number of smaller firms. The analogies through which people describe how DNA synthesis fits into the emerging bioeconomy tend to be based on thinking about other industries, and it may be that not all those parallels are useful.

Not every lesson from chips or software is transferable to biology.

You sound optimistic about new firms entering the market…?

We’re still living the story and it’s hardly set history, so it’s important to articulate all these things as hypotheses. Many industries wind up being dominated by winner-take-all structures or by firms that engineer the market that way. But I suspect biology isn’t going to go this route. In 2000, when I started keeping track of DNA synthesis, sequencing, and productivity, and then trying to compare that data to other industries, it wasn’t clear that Intel would be the sole winner. It was still useful to ask what we could learn from chips and transistors that might help understand where biology was going. A chip fab then cost about USD $1 billion, which was expensive but lots of people were trying to play. Today, it costs USD $10 billion for a cutting-edge chip fab. Very few firms can afford that much capital expenditure, especially when it’s not clear how big the market will be. The analogy between biology and chips is now more instructive for the differences than for the similarities.

In contrast to the increasing costs for chips, if you look at synthetic biology and biotechnology broadly, nothing costs USD $1 billion. Even developing a drug doesn’t cost as much as companies say it does because they’re amortizing over all of their failures. And, in general, it looks like costs across biotech are falling. Anything to do with reading and writing DNA, and with protein expression and structure finding, is falling exponentially in cost. My sense is that the cost of complex metabolic engineering is also falling steeply, although there is limited data to back that up. Still, for a biochemical or other industrial biotech product, a company might now spend $100 million or less, soup to nuts, getting to market. Basically, the barriers to entry to do biology are lower than other industries. That will lead biology in new directions.

When you think about our industry through the lens of the synthetic biology stack, what parts are most interesting or lucrative to you these days?

Well, I like the whole stack in the sense that it all has to work in order for us to get to true bioengineering, and true design for manufacturing for biology. But you do have to pick your shots. We have experience as engineers and entrepreneurs. We know how to build and ship products, what it means to make and sell a thing, how much it costs and why. When we are looking for technologies to invest in, we look at where the pain points are, what’s hard, what’s been invested in already, and what’s under-invested in. Where can we apply a lever to move the whole industry forward, and make money for our investors at the same time?

When you look at the cost of getting a product market today, most of that is in scale up, which is manufacturing development and process development. That has multiple components, but it includes carefully laying out the problem you think that you’re solving. How do you get all the way from the whiteboard or napkin to a finished product? If you don’t do that in a thoughtful way in biology, you can wind up innovating yourself into a corner with something that sounds good, looks good, and that you imagine people will buy, but that turns out to be very hard to manufacture.

In other industries, that problem is solved with a software stack that starts with simulation, which then is connected to the ability to do experiments in a rapid, closed-loop fashion to avoid getting stuck in that corner. We need to build that capability for biology. As an investment firm, we’re not particularly focused on hardware or software or wetware, but again we are looking for where the cost is. Where is the hardest thing, the hardest problem that has to get solved? Where can we bring down the total cost by injecting capital and through some thoughtful development of technology?

Bioeconomy

The US bioeconomy reached $370B in 2016, according to the Bioeconomy Dashboard. One-third of that is the industrial biotech sector, which is growing faster than crops or drugs on its way to outperforming petroleum in the chemicals sector. Courtesy of Rob Carlson, Bioeconomy Capital.

Your venture fund, Bioeconomy Capital, invests in companies that develop enabling technologies, and companies that use those technologies to make things. What are the differential challenges of investing in those two flavors of companies?

We choose our investments based on our experience in engineering, manufacturing, and strategy. Our investing goes all the way from experimental layout and planning tools through to manufacturing tools, and then linking them all together. We don’t imagine that we know specifically what all these technologies are, but we think we know what they need to do, broadly speaking.

Say you are developing a new tool. We try to help companies get through that process and then find a market. That is a hard problem in a sector like pharma, where you have a two-year sales cycle and huge effort just to get in the door. Once you make your pitch, then you have to grapple with the fact that they are on a yearly budgeting cycle, and if you don’t get your big sale on the budget and approved by the CFO in September or October, whether it’s software, hardware, or some other big expenditure, then you’re not getting on the budget for another year. A $100,000 proof-of-concept project can be approved anytime. But a $5 million tools order can take years to get finalized, and that sales cycle is extremely hard for startups to navigate. A handful of proof-of-concept projects looks good for your Seed or A Round, but it takes real sales to grow a company and to show traction for the B and C Rounds. If your customers all have long sales cycles, it can take a long time to get to significant revenues. That is one reason why we are careful about investing in startups that are targeting the pharma industry for sales. The pharma industry recognizes that it needs to implement radical change, but it also has trouble moving quickly. We expect to get more traction from large chemical companies, and companies wanting to engineer organisms to make chemicals, because those organizations iterate faster and they’re not quite as bound up in pharma’s ways of doing things. That transition is coming, but it, too, can be slow.

Then on the products side, you may have a company that has already developed a product, say a drop-in replacement for a petroleum-based pigment that a big customer is clamoring for, and your problem is how to make enough of it to become commercially relevant and survive. It poses a completely different set of engineering and process challenges: It’s less about sales and more about implementing manufacturing and getting whatever that thing is out the door in volume. It’s a lot of hard work in both cases, but it’s a different kind of work.

What do you think of the wisdom that ‘investors invest in people, not things’?

The old adage that we go by is “team, technology, and market”, but team is always going to be first on the list. We might feel like the market is nascent, and as long as the technology looks really good we can hang on a while and let the market emerge. Or we might feel like there’s clearly a market for this product and we believe this team can develop whatever the widget is and fulfill that market demand. But you have to have people that can execute. That could mean just following the plan that everyone agrees on in the beginning, or it could mean finding people who are flexible enough to recognize that either the plan needs to be re-done, or that they are not the right people to lead that plan. That’s a hard thing for some people, because even if you think you know more about the technology and how it fits into things than anyone else in the world, you might need serious help on the executive side, or on finance, or in operations. You have to be flexible about bringing in other team members because without the right people you can’t do anything.

How did your experience in the garage bio scene inform your investment perspective?

Besides the garage startup experience, I had many years in academia, and I continue to consult on other commercial projects in part to maintain my technical chops and my touch on how things are changing. All of that feeds into decisions about what kind of tools the market needs. Even if we can’t identify a specific biological tool, we know we need a kind of wrench, or a kind of soldering iron, or a multichannel oscilloscope. We take our industry experience and use it to identify the capabilities of the tools we need to invest in next.

The garage specifically taught me about getting by with minimal resources. I got a lot done in my garage. For example, I developed a technology to quantify 1,000 different proteins from a small sample at once. It worked, and it was commercializable. But it’s a technology that turned out to be at least 10 years too early, and now I think maybe 20 years too early. Today we receive proposals to do almost exactly the same thing that I was doing in my garage 10-15 years ago. And I still think it’s too early. In this specific case, I don’t think there are enough customers out there who recognize the need to quantify 1,000 proteins from a small sample at once. Sure, you can find a couple of customers who pipe up and say this sounds cool. But you need 500 customers like that to make a product fly. If no one buys it, knows they need it, or can fit it into their R&D pipeline, then it’s not that useful. That was a hugely important [garage bio] lesson — painful, to be sure — but it definitely impacted the way I think about investing.

What trends in the bioeconomy are tickling your curiosity these days?

I continue to be struck by how much faster industrial biotech is growing than either drugs or crops. The data are terrible, but biochemicals appear to already make up one-sixth of total chemical sales in the US, or more than USD $100 billion in 2016. That is shockingly large, and it could be wrong. But the more data I see, the more I believe that it’s true. And that market is growing quite fast — maybe 15% a year. So what I’m excited about is that as biology increasingly outcompetes petrochemicals based on price and performance, there is potential for huge change in the world.

Even amidst the massive hype 10-15 years ago, the hypothesis was that maybe biofuels weren’t the right way to pursue this whole program. Fuels are high volume and low margin, whereas maybe you should look at the other end of the barrel to get the most traction. I always had my doubts that biofuels would be important in the near- to medium-term. A key hypothesis for Bioeconomy Fund 1 was that biochemicals, which can be low volume and high margin, would start to take off as both engineering and manufacturing matured. It has become quite clear that this hypothesis is correct.

Next, it is important to understand that biology isn’t the only fast moving technology in the world. Solar power and batteries are now making huge strides, and transportation is being electrified much faster than anybody expected. So the demand for petroleum — for either fuels or chemicals — is being attacked from both ends of the barrel. The demand for petroleum is going to peak sooner than expected. The IEA forecasts that the largest new demand for oil over the next decade is going to be in petrochemicals and plastics, but this is exactly where biology is making big strides.

Given that it costs USD $20-$30 billion to build a petrochemical facility, and a biological manufacturing facility looks a lot more like a beer brewery, with lower costs, then biology is going to win that battle.

We can already make everything found in a barrel of petroleum in real time using engineered organisms. The capital costs are lower, the timescales to get the manufacturing spun up and working are lower. Biology isn’t universally outcompeting petrochemicals yet, but it’s clear that that’s coming. The petroleum industry, which sees its largest growth coming from one particular sub sector — chemicals — is going to be disappointed.

Petroleum is not going to die by anyone immediately pumping out large volumes of either biofuels or chemicals, [but rather] it’s going to die because the financing stops making sense. Petroleum mining and processing infrastructure isn’t merely expensive to build, it is terribly expensive to maintain. As demand peaks and then starts to decline, banks are going to stop lending to the industry, maintenance will be deferred, and insurers are going to stop insuring both the facilities and the debt. It is simply a matter of risk and return on capital.

The combination of biology and electrification are going to dramatically reduce demand for petroleum, which is eventually going to make it a bad investment. By 2025, I think we’re going to see big impacts, and that’s exciting.

How has your thinking evolved since publishing Biology is Technology back in 2010?

There’s something that I didn’t grapple with in the book, because there was no data to grapple with at the time, and that’s the idea that better measurement improves predictability. For the last hundred years, we’ve told ourselves that biology can’t be quantified, it’s too squishy, it can’t be engineered, that’s not the way biology works — you just can’t do it. But it turns out that we were just bad at doing biology. As we bring down the error bars on measurement, we can use that improved data precision to drive modeling and engineering. If you look out the window, it’s pretty clear that biology is entirely predictable. Biochemistry is entirely predictable, once you get the right physics written down. Organisms are the same every day: yeast is yeast and E. coli is E. coli. They don’t change much, and we can increasingly predict these changes. Biological engineering is finally beginning to look predictable, too. There are a number of examples in the literature in which driving down measurement error bars leads to dramatic improvement in the ability to engineer biological systems, which then leads to predictable manufacturing process development. We can now predict with >90% accuracy what happens when we transfer engineered organisms from small lab volumes to 1,000- or 100,000- liter manufacturing volumes. And we will get even better as new tools and process development technology comes into play. That is going to make product development a lot easier. It’s going to make the economic shift from the old way of manufacturing things to biological manufacturing that much faster.

Another thing I’m pondering is the importance of the intersections of what might look like completely different technologies. Biology isn’t the only story in town. Taking the earlier example, biofuels are economically and technically extremely hard to pull off at large scale, because you have to be enormous in order to make enough money on a low-margin commodity to be able to run a business. A biological technology has to work really well to beat fossil fuels, which has proven more difficult than many expected. So maybe that is not the right problem to try to solve. The government could, by fiat — gas taxes, carbon taxes, etc — try to make fossil fuels more expensive, but that’s proved difficult politically. Perhaps it’s better to just win economically. New, unsubsidized installations of renewable power generation and batteries are now becoming cheaper than running existing fossil fuel generating capacity. Electric vehicles are falling steeply in price while improving in performance, and consequently people are buying them. The rapid roll out of electric buses is putting a noticeable dent in Chinese diesel demand. But all that electric power production and usage isn’t going to help replace petrochemicals, that is, the atoms that we build our world out of.

Biology is starting to make real inroads into chemicals production, and will eventually outcompete petrochemicals on both price and performance. The combination of those economic and technical trends is going to change the world in dramatic ways over the next ten years.

Kevin Costa

Kevin Costa

As Editor and Program Manager, Kevin leads SynBioBeta's digital media content and works with customers and partners to build a world-class community of innovators. Before joining SynBioBeta, Kevin managed the Synthetic Biology Engineering Research Center. His interests include public engagement, science writing, community building, and bikes!

Stay updated on the latest news in synthetic biology

Join our weekly newsletter

Sign up

Job opportunities

More