Great analysis, Lee. I think your LTV/CAC math is overly optimistic, though, as it doesn’t factor in time value of money.

Let’s assume just a 20% cost of capital (I usually use 30% for startups but BlueApron is more mature, so 20% seems right — mature public companies cost of capital is more like 8–12%. Then, if you discount each of the three years of revenue by 20% (compounded appropriately for years 2 and 3), you get an LTV of $680. Margins of 31% suggest the net contribution is more like $211, or roughly 2x LTV/CAC. Still solid, but not nearly as attractive. If you believe the next wave of customers will cost more (reference soaring marketing costs), because customer acquisition gets worse with scale, there could be some real challenges ahead to make the unit economics work.