Initial Non-Recurring Recurring Launcher $110 0 110  Carrier Spacecraft 50 30 20  Robotic Element 70 50 20  Mission Control Element 40 35 5  Recovery Operations 30 2 28  Marketing 50 40 10  Program Wraps 70 32 38  Totals 420 189 231 
 Launch costs: I got that launch cost from the launcher data we have on line in section 220.127.116.11.1. of the Artemis Data Book. That's the cost for a Titan 3 launch. We're assuming we'll want a fairly heavy payload in Low Earth Orbit (LEO).
A Titan 3 will put 26,000 lbs (11,700 kg) into LEO. Determining how much lunar sample payload we can return to Earth with that launch mass depends on what we assume about our spacecraft. The big drivers are the type of rocket we use (hypergolic vs cryogenic propellants) and what assumptions we make about the weight of a heat shield. We can assume a direct entry on return from the moon, since our payload (rocks) doesn't care if we accelerate it to 14 g's to bring it home through Earth's atmosphere.
Delta III, Proton, or Ariane 5 might save us a some launch costs, but at least this gives us a number to start with. We need to research alternative launchers, such as the Delta III.
About units: That comparison-table document breaks a rule since it only gives payload mass in French Revolution units. The conversion factor is 2.2046226 lbs/kg. This is another plea for everyone to follow the program's convention of quoting units first in the English system, then in the metric system. A couple of minutes of your time with a calculator when you produce a document will save thousands of hours for everyone using your data. In this case, it would have taken only 30 seconds in a spreadsheet to fix it.
 Carrier Spacecraft: This is the vehicle that carries the surface equipment to the moon and then home again. I guessed about 20% of development costs for the Artemis spacecraft. It sounds about right, based on no human element, very relaxed safety requirements, commercial approach. About 60% of total cost would be for engineering, which we only have to do once; so those are non-recurring costs if we want to do this more than once.
 Robotic Element: About the cost of two rovers, FOB Carnegie-Mellon University. A machine that can run around and pick up rocks, and then ship them home, seems to be about twice as complex as as the Lunar Rover Initiative robot. I assumed we'd be able to reuse most of the robotic element for subsequent sample return flights.
 Mission Control Element: This is a rather complex facility, with tracking and communication stations all around the world as well as a central mission control facility. People will have to plan and execute the mission. We'll need an environmentally controlled laboratory for sample processing.
 Recovery Operations: This is a really wild guess. This assumes a water recovery, where we lease a small ocean craft to snag the package out of the water and then lease an aircraft to carry it home.
 Marketing costs: Another wild guess. We'd do some direct mail, but mostly this is for television advertising. Assume subsequent missions would only need about 20% more marketing to have the same impact once the initial messages were developed and the public familiarized with the product. Microsoft spent about $200 million on Win95's marketing campaign, so this seems about right.
Let's also assume we'll photograph each sample and put a note on the web site about who owns it in an on-line Registry of Lunar Samples. (Of course, buyers will have the right to privacy, if they prefer.)
 Program Wraps: I added a 20% factor to account for management, lawyers, business office, paper clips, and all those fiddly things.
 Totals: So, for a total investment of $420 million, we could have a nice pile of rocks and dust in an environmentally controlled warehouse, and a stack of orders which presumably equals our stock of trade goods.
Let's assume that packaging and shipping of the samples is negligible, since we could break it up into 10,000 samples with $500 each for packaging ($5 million total) and still be well within the noise level of these estimates.
Note that the recurring cost estimate says we could do it again for an additional $231 million.
 Initial Cost: This is what it will cost for the first mission.
 Non-recurring cost: This is how much of the initial cost represents a permanent investment. This value is carried over if we do more than one lunar sample recovery mission.
 Recurring cost: This is how much it will cost us to do it again, once we've done it the first time.