The main problem was what would you expect this to do.
One could send a very small impactor - say, a fraction of a gram of tungsten.
Send a 3U cubesat to LEO, equipped with a ion drive the kind Busek manufactures. Using DSN guide it into impact trajectory of Mars. All contact is lost a couple hours later, and while there are some inaccuracies, gravity of Mars will be enough to finish the "guidance process" for you. The cubesat burns up in the atmosphere, but the tiny slug wrapped in a ball of ablator impacts the surface, and evaporates on impact, with no consequence whatsoever and in a way that is entirely unobservable from Earth. Nobody is any wiser and there's absolutely no point in this whole endeavor. Meanwhile, the whole endeavor costs a couple million dollars.
Is that the effect you wanted to get?
If you want something marginally useful, the scale grows immensely. Instead of the (easy) impact trajectory, you need an orbital capture trajectory, likely with aerocapture; the difference between shooting a barn and shooting the roof of a barn in such a way the bullet passes through, but lands in the gullet on the other side. And still, trying to go "tiny" means you won't have room for any usable payload, so the endeavor's purpose is dubious. The gap between "a mission to Mars that makes sense", and "having some atoms from Earth infuse Martian soil" is huge, and we have no clue what exactly you expect.
Edit: Fair point, Michael.
I think the right answer to the general meta question here would be:
Yes - you'll get good answers only to good technical questions. But that doesn't mean you can't ask a broad, general, misinformed or otherwise "bad" questions. Just ask for assistance, and we'll help you reshape them into good questions, and then provide good answers. You're not expected to know how to phrase your question to be perfect right off the bat - but you're expected to be willing to work on improving it, as its shortcomings are becoming apparent to you.
We can't give you a good, precise answer if there are dozens of valid answers depending on conditions you didn't detail. And you are well within your right not to realize these conditions exist at all. The 'close' as 'Too Broad' shouldn't be the end of life of a question - it should be the beginning of the cycle of its improvement. Asking what is it that makes the question too broad, getting the answer, then editing the question.
Even if you still aren't sure which option to choose, or how to narrow it down, you can still edit the question to ask for an overview of the possibilities. Instead of one, precise, in-depth solution, you'll get an explanation of how the problem and solutions vary with changing parameters. Then you may re-ask another question, which focuses precisely on one section of that, the one you're actually interested in.