# What does it mean for a technology to scale? People throw around the question “does this technology scale?” a lot. We rarely unpack what that means and (I suspect) many of us don’t even know. Scalability means many things to many people but it’s worth unpacking because ultimately There’s a narrow definition of scaling that asks “what happens to the price of a thing when you make more of it?” Obviously there’s no clear boundary and scalability is [[Nebulous]] but it’s worth trying to unpack what it means for a technology to scale. Scalability revolves around the simple question: **“can this process make enough of a thing at a price that people want it?”** It means different things for a drug manufacturing process to scale than for a commodity chemical manufacturing process to scale because consumers are both willing to pay far more for drugs and need far smaller volumes for them to be effective. Cancer drugs can cost ~$10,000 for a ~100mg dose; ammonia costs ~$1500 for a metric ton which can fertilize ~10 acres of corn. Perhaps the best way to dig deeper into what it means for a technology to scale is to upack failure to scale. There are two ways that a process can fail to scale: 1. It can have a “hard” cap imposed by some external bottleneck 2. It can have a “soft” cap created by a too-high price at the volumes necessary for a product to be useful #### Hard-cap scaling failure It wouldn’t matter if you created a new process that could create ammonia at $150/metric ton if it could only produce a single ton of it at that price. But can’t you just do that single-ton process many times? If you can, you can either scale or enter soft-cap scaling failure. However, some processes have hard limits imposed by an external bottleneck. This bottleneck could be on the capital infrastructure to parallelize the process like some specialized piece of equipment that is only produced at a certain rate (because of its own scaling constraints). A bottleneck could also be on an input to the process: maybe it requires some rare material that is only produced at a certain rate. #### Soft-cap scaling failure Most scaling failures aren’t hard limits imposed by external constraints, but instead situations where making enough of something to be useful would cost more than its consumers are willing to pay for it. To talk about cost-related scalability, we need to unpack what goes into the cost of a process: capital expenses and operational expenses. Capital expenses (capex) are the costs to set up a process: equipment, factory real estate, infrastructure, etc. Operational expenses (opex) are recurring costs that need to be paid every time you run the process: raw inputs, labor, waste disposal, etc. Capex factors into the price of a process ==here== Another question around scalability is whether you can maintain a linear or supra-linear relationship between inputs and outputs. That is, **how much money/work do you need to spend beyond the costs of raw inputs to increase outputs?** A process can scale not by decreasing the cost of the output, but by figuring out ways to sell byproducts An implicit part of making a thing at a price that people is the A perfectly scalable process requires no ongoing costs beyond **Does the product maintain or improve in qualities when you make more of it?** ### Related - [[You need a certain scale for technology development to be worthwhile]]