A while back, I talked about my novel submission coming back from Baen. I submitted it using their online submission process. I had already been waiting about ten months when I visited Alan Dean Foster's home. We discussed it for a few minutes, and he told me that he and his agent recently had discussions with Baen, and that Baen was only buying alt history and military SF. Since my novel is neither, I knew the manuscript would be rejected, it was just a matter of how long I had to wait.
I finally got the rejection in October. I decided to try the agent route instead of over-the-transom. My SFWA membership is worth the price for the SFWA directory alone. It lists who represents who, a priceless resource for somebody looking for an agent. You want to find somebody who represents similar writers, and you want a good filter to screen out the crooks. Chances are pretty good that if well-known-writer X is represented by agent Y, it's a legitimate operation.
Preparing the query package takes attention to detail. Every agent wants something different, but there is also a lot of overlap. For this novel, Neanderthal Swan Song, I prepared around a dozen files so I mostly have to just grab the ones I need and print them out. If I come across a new file, I'll make it and park it in the parent directory.
I then make a subdirectory for each agent an drop into that folder everything I send that agent's way. Most of it is mindless, but the cover letter (aka query letter) is not. Go look at this agent's blog, especially this entry. (For the record, I have not queried him.) Note the point about it being bad form to let him know who you queried before him.
Everyone knows you don't rewrite the cover letter from scratch, not since they invented the hard drive, anyway. Everyone recycles letters like this, but you have to go over that cover letter very carefully to ensure you spell the names right, you get the right enclosures listed, and correct any other legacy points that came from the last agent you queried.
So anyway, as I watched the Cardinals post 9-7 as their 2008 season record, I went over a list of 27 agents that I culled from the SFWA directory based on having a New York proximity (with one exception) and having at least one client with a name I recognized. I narrowed it down to my top picks and prepared 7 query packages this evening.
This business is a numbers game. Read about that aspect from a guy with 100 novels, Dean Wesley Smith. It took me a long time to start finding my niche markets with short fiction. I submitted to all the same markets that my writer friends sold to, but I never sold there because I don't write what those markets want. I write what different markets want. It's not a statement about whether that is better or worse, it's just different. The same goes for agents.
I won't repeat the stories of the writer whose manuscript was rejected 50 times then made millions of dollars on the 51st try. But it does happen. If you are competent, there is a market waiting for you. It's just a matter of trying to find it. That's a numbers game.
I sold 7 short stories in 2008. I was disappointed with my year (mostly due to all the distractions in my life) but another writer commented that he would be delighted to sell 7 stories in one year. But, I also collected almost 100 rejections to sell those 7 stories. Not everyone is Jay Lake, who claims to sell about 1 in 3 submissions. I'm not, but then, he's been at this game longer than I have. Maybe someday I will, too.
But, that isn't the point. The point is that you can't read an agent's mind from across the room, much less from 2000 miles away. So it has to be a numbers game. You don't know which agent will like your work enough to want to work for you. That means you have to find the ones you want to work with and work your way down the list, just like with marketing short fiction.