I base this on the assumption that we are aiming for having any kind of plant life on planets far away that do not currently harbour life (not necessarily spreading Earth's life).
To describe an upper bound of the method, I'll look at it from a high level perspective, refering to the other answers for the "technical details".
Based on this, I will show that - depending on various unknown estimates - the method of sending germs may not be worth it - because nature is faster:
The idea is doing that by spreading life from earth to other planets, using some kind of suitable germs. The germs do not use propulsion, and are send out in random directions.
Based on other answers, I assume it would require a very huge, even astronomical number of germs:
So, we hope to create one or more plants during the next couple of billion years.
We need a large number of germs to expect even a single one point at a suitable planet. But very few of these will end up on the surface of the planet and be still alive (and that's already assuming they stay fresh for a billions of years.)
All in all, that will be expensive. But let's assume we have "suitable funding".
So, we can plan to use all resources of the earth to produce and send germs.
Now we could, in principle, calculate a probability of succeeding, assuming we go to that limit.
Now, the interesting point: I suspect that it may be more probable that live develops by independent evolution on the planets.
Of course, the probability of life starting from molecules is hard to tell, but, for some definition of "succeeding" above, we could compare to find out whether nature may succed way before us.