Good idea:
Searching Twitter for mentions of your product or service or brand, and conducting direct Twitter outreach to stave off a crisis or solve a problem. Comcast does this famously well; other companies are learning.
Last month, I vented about a string of catastrophic outages by our microcephalic e-mail hosting provider. By the next day, a representative had found the Tweet and contacted me:

we still fired him — but thanks for playing!
Inscrutable idea:
Searching Twitter for keywords vaguely associated with your product or service or brand, and indiscriminately re-Tweeting them.
Example:
Last night, I Tweeted about eating in bed (always a bad idea). My friend made a joke about bedbugs (okay). Within hours, a bedbug expert (who?) picked up my friend’s Tweet, and automagically re-ran it … along with a link back to my friend’s Tweet:

"I have nothing to contribute to this conversation"
That’s it? Really?
Bedbug Guy didn’t then then link to a … blog about bedbugs? Or offer some sort of commentary with which to define himself as the bedbug expert? Or do anything in the way of promotion or outreach or contribution to the bedbug conversation?
Nope.
So this doesn’t even count as Twitter spam per se, just meaningless repetition. I just don’t see the point.
Bedbug Guy: I never thought I’d have to give this advice to anyone, but — if you want to effectively pollute Twitter with mindless spam, you’re going to have to try harder than this.
Some people.