In summary, we’re pleased with the overall direction of progress and continue to focus on scaling the business as we finish out the year. We remain focused on driving new programs to the platform in this challenging macro environment. We’re excited about the significant traction we have made in the biopharma segment. And we continue to manage our balance sheet and cash flows to maintain a long runway while maintaining flexibility to capitalize on near-term strategic opportunities, with over $1 billion of liquidity at quarter end. And now, Jason, back to you.
Jason Kelly: Thanks, Mark. This is a solid quarter for Ginkgo. Our deal with Google sets us up well to lead in the application of AI to design DNA and proteins. While our deal with Pfizer is a real signal of commercial progress that I’m going to be digging in on a second. So however, I want to address why we’re taking down cell engineering guidance. We’re building our relationship with you all as a young public company. And so while we want to have ambitious but achievable goals, we also want to update them as the year progresses and tighten ranges as we get close to year-end. So we’re revising our guidance on the cell engineering services components of our revenue to $140 million to $145 million, down from $145 million to $160 million.
Generally, this is for the reasons I provided on the last call around industrial biotech venture capital drying up, and also — and reducing the size of programs that we’re seeing in that sector as well as our new program counts being lower than hoped for in Q3, which impacts Q4 revenue. Now I do want to spend some time on the program counts being lower because this is a critical metric for us that demonstrates our flywheel spinning up at Ginkgo, we get better with scale. And it’s one that I pay a lot of attention to internally here. So we had 21 new programs this quarter, which was less than I hoped to get. But at the same time, our enterprise sales infrastructure is stronger than it’s ever been at Ginkgo. And in particular, I want to call out our new program with Pfizer and explain why it’s an important demonstration of our commercial capabilities here.
So this is a drug discovery deal in mRNA therapeutics. And that’s important, first, because drug discovery is a harder sell than manufacturing R&D deals, which is a number of the previous deals we’ve done in biopharma. And the reason for that is discovery work is more closely held by our customers. Remember, we have to convince customers to outsource work to Ginkgo’s platform that they would otherwise do themselves. This is the kind of work that they tend to think they should do themselves, all right? And then secondly, mRNA is a new modality. It’s a new type of drug, right? And so it’s emerging, it’s high tech. And Ginkgo is proving that we can lead, right, because customers have to choose to work with us in an area like mRNA as a general platform.
In other words, the same platform that’s doing mRNA biotech is doing agricultural biotech. And all this is to say that this is not an easy deal to close, especially with several hundred millions of dollars of potential downstream value attached to it, which is why it’s worth pointing out that I wasn’t involved heavily in closing this deal nor was Jen Wipf who heads up our commercial team. This really came out of a normal sales process from our commercial and deal teams here at Ginkgo. And this is a big deal because as much as I like to think myself as good at sales, I’m not scalable, okay? Jen is not scalable. Our enterprise sales team is scalable, and the types of deals Ginkgo is doing, involving fees during technical work, plus hundreds of millions of dollars downstream milestones or royalties, are typically negotiated by the CEO and leadership team of a small biotech company if they were partnering with a large biopharma like Pfizer, like that type of deal you see popping up all the time in industry press and so on in the pharma industry.
Ginkgo being able to do a deal like that in a routine manner is a huge strategic advantage for the company and the result of great work and team building, led by Jen, who heads our commercial team over the last 2 years to really build an enterprise sales engine here at Ginkgo. So I’m thrilled to see that. And if we hit the high end of our updated guidance of 85 programs, that would work out to 30 new program starts in Q4, which would be a great further signal of how we’re scaling this enterprise sales infrastructure, and that’s something I’ll be watching with the commercial team coming up. Finally, with the updated guidance, we’re still looking at 36% to 44% growth in new programs and 32% to 37% growth in cell engineering services revenue over the last year.
Scale helps Ginkgo, and so I’m happy to see that rate of growth. Okay. So let’s dive in on our 3 strategic topics. So first, I want to share some recent customer case studies, where we apply our AI technology, so you get a little more sense of what we’re doing there. Second, we often get asked, actually, get asked a lot about what programs we’re most excited about and what programs are most advanced, sort of our program pipeline. I’m not going to pick favorites. But I’m going to share a lot more data around that pipeline, so you get a better understanding of where all those active programs are and just how much diversity there is in those programs. And then finally, I want to share a little more about how we’re thinking about the future of our biosecurity business as a defense technology business, as part of national security infrastructure, and how importantly it relates back to cell engineering.
Okay. Let’s jump in. Okay. So first, I want to talk about how AI fits into the other assets of Ginkgo and show some case studies of its application. So we talked a lot previously about our foundry and code base at Ginkgo. A reminder, our foundry is our automated laboratories here in Boston that generate data at lower costs as they grow in scale, okay? Think like a factory for testing genetic designs. This data is organized into what we call our code base, which is reused across many different customer programs. In other words, we can use data from one project to help speed the development of a second project from a different customer. And this is, again, another important asset that gets better at scale. So what’s exciting is that these data assets can also be used to train large AI models that then inform the sort of experiments we should do to better train those very models, okay?
