Having tried a couple of times and basically failed, I can list a bunch of points about what can go wrong. But that isn't the same thing as knowing how to do it right of course.

I have basically given up, except for the occasional one-off, and if your blog is like mine, you may want to consider giving up as well. Guest-blogging programs are not for every blog.

I have had maybe half a dozen guest posts on my blog. Several interested people have started, but failed to finish, guest pieces. I get a lot of splog guest posting requests from Demand-media type people who want to post SEOed pieces about online MBA programs. But so far, I've not had a single "Wow! This guy wants to write for me?" in-bound walk-in. I have solicited a few, but not systematically, and my hit rate is pretty low, like 1 in 5 or so, of the people I ask actually turning a piece in. Another 1 in 5 informally commits and sends me a draft/idea or something, but never finishes. The hit rate is not enough for me to put in more effort.

The few pieces that have made it past the barriers haven't really been extreme successes, either for the guest or for me. No viral hits, but no embarrassing flops either.

Final piece of data: I have around 2500 RSS subscribers right now (Jan 2011), an email list of about 700, maybe 500-1000 visits a day, average. So... a decent C-list blog, I'd say. On paper, well worth the exposure for D-listers. So why haven't I successfully launched a guest program, when many blogs half my size manage it?

Analysis.

Your blog's brand has to be "guest friendly." I didn't realize how closely my blog's brand was associated with my own writing style. My readers have on occasion been vicious and ruthless in their comments on guest posts, in ways they never are with my own posts. One commenter bluntly posted something like, "this is crap, I come here for your writing. If you plan to slack off by inviting bad guest posters, I am going to stop." The guest post in question wasn't bad, just not in my style. A strong brand creates a rejection immune system in your own audience. This is strong enough that nowadays, I tell well-intentioned potential contributors: "your piece is good, and you may think guest posting here will help you, but it will actually hurt you, so no, thanks, for your own good."

Basically, the more unique you are, the less replaceable you are. For this reason, some savvy bloggers avoid becoming too unique, and maintain a commodity streak to their brand. Keeps their platform accessible to guests. Such bloggers usually also have a quantity-over-quality model, and post every day, sometimes 2-3 times a day, and look for regular guest posters to take over the weekend slots. The barrier to entry is low, since the frequency implies short length: 300-500 words will do, and there is more room for error due to the high frequency: a single bad post won't hurt such a blog much.

If you plan to "flip" a blog (grow it and sell it) make sure you don't make the brand too strong or too closely associated with yourself. Keep it somewhat within a commodity niche, and initiate a guesting program from Day 1, to divorce the blog voice from your personal voice.

For me, it probably doesn't help that I only post once a week (so... 52 a year; which a "speed" blog will crank out in less than a month), and have created strong voice-branded quality-over-quantity expectations (my readers now seem to expect a certain "ribbonfarmesque" voice to all posts, which is narrower than my own natural voice, that even I cannot risk violating).

It also doesn't help that my blog has no clear niche/theme. It is basically in the "eclectic" category, which means it is in no category. Which means it ends up being about (a narrow aspect of) me by default, even though it isn't a "personal" blog (I am not personally interesting enough to sustain a blog about myself).

Which means I have very few open slots, and demand significantly higher quality, post length and "on-brand voice" constraints for those slots, which makes it not worthwhile for good writers, given my quality/reach ratio. Basically my barrier to entry is far too high for any potential value somebody could get out of it.

My only hope is really good, but completely unknown starting bloggers. I have no idea how to find those. I'll get a week off, but no real immediate exposure benefits, and the newbie will get a launch boost, and end up in a position where more posts on my blog won't help them.

So I basically need a steady supply of highly-undervalued newbies, who need a one-off launch boost.