All right. So let me give you a little bit of the historical numbers, and how that has looked for us in the
past. So on the foundry revenue side, this is pretty straightforward. And so, one of the things I would highlight thats interesting, is if you look at the cost to do a cell programming project at Ginkgo, back in 2014, the deals I did in
14 and 15
100% full-burden cost... We were, from customers, getting less than 20% of the costs being paid in revenue from the customers, so Ginkgo was bearing 80% of the cost. Now, why was that? Well, I was asking for intellectual
property rights to go into my Codebase, and also royalty or equity as a value share on the backside. And my platform, going back four years, substantially worse than it is today. Ive been basically doubling the efficiency of the platform
annually over that period of time. In fact, if you look back on that slide where I showed the numbers, I had a dotted line that showed horizontally the cost for a scientist at the bench doing this work today.
And today, we beat that scientist, about 5-10X cheaper, when we benchmarked the work. Back in 2014, 15, I was actually more expensive. So Id be
more expensive than your own R&D team doing this. And so, you can see we held the business model, but we gave on the economics on Foundry. Now, whats happened over the last six years, is as our platform has improved, weve been able
to capture more of that foundry revenue spend of the cost of a program covered by the customers, to the point where last year, about 80% of the program cost was paid by the customer, and we expect as our costs continue to fall, we can start to
actually make a margin on that foundry revenue in the future while still retaining those intellectual property rights, and that royalty or equity value share. All right. And so, thats the left side, and the great thing it is, its pretty
predictable.
We are doing these deals, theyre often one to three year contracts, so we have good visibility. The majority of my revenue in 2021 on
the foundry revenue side was actually booked prior to the year starting, so I like that as a public market metric for us. On the downstream value, theres two ways we can talk about this. One is predictable, and one is unpredictable. I will cut
to the chase and say we went with the predictable one. But to give you a sense of our choice on what is unpredictable is, when is, say the project with Roche going to complete and give us a royalty on that drug sales? Who knows, right? It could be a
year out, it could be five years out, theres a lot of variability in that process. Which programs, of all the programs ongoing at Ginkgo are going to be the ones that technically succeed, and which ones wont? Also unpredictable.
Whats predictable, on the other hand, is how many new programs will I do in the next quarter. That six-month
sales cycle with these large customers, we have good visibility into how many new programs were going to do. We did seven new programs just in the first quarter of this year. Four were inside sales, three were new customers. We have good
visibility into those 23 Im trying to hit this year. I feel good about our ability to add new programs. And then what I can also say is, All right, what has a program been worth historically to Ginkgo? So that as we add a new program,
investors could give us credit for that implied value of that new program, regardless of which ones going to be successful? And so, if you look at the last fifth... Since 2017, Ive done 54 deals that we either have a royalty, or in
lieu of a royalty, equity in the partner company. And so the royalty ones, the challenge is they take a while.
I dont get a royalty until not only
have I completed the technical work, but the customer commercializes that product, and then I get a value back on the sales. All right. So only some of our earliest programs and things like flavors and fragrances give us royalties today, we
dont have great data
Page 12 of 15