I've left unpublished a bunch of comments lately that seem to reflect a new kind of spam.
Since Google's algorithm is known to value inbound links, many of the spam comments that get posted on the Web actually aren't intended for users of the site where they're posted. They're Google spider bait. And when they're posted on a website that has good Google karma -- like mine -- they can help elevate the target site in Google search returns.
Until recently, the comment spam that I routinely delete without publication has been heavy-handed, obvious, and probably automated.
But now I'm seeing spam that includes a quote from the blog item and some innocuous question like "Is this practical?" or a meaningless comment like "All of the above? Could be." And the link to the spammer's website is subtly tied to the username.
I may respond to this by simply disallowing links of any kind, but that defeats the purpose of linking to legitimate blogs from legitimate comments. For the time being I'm going to simply be very suspicious of any comments that don't pass the "sniff test."
Comments
reading comment threads becomes a guessing game.
I've come across those kind of comments and thought I was missing some kind of inside baseball. Its really irritating as a reader to have half-connected comments that are enough connected that you don't immediately dismiss them, but not connected enough to know what they mean.
When in doubt
I always follow the "when in doubt, delete" approach. Because if the person is making a comment so useless that it leaves you wondering if it's spam, then chances are deleting the comment entirely won't be a huge loss.