I've known Andrew for a very long time. He was one of the last young publishing up-and-comers to cut his teeth working for Donald I. Fine, and for some categories of work I'd unhesitatingly suggest him; I quite envy in particular the breadth of his non-fiction offerings. But I don't agree with his assessment of Bookscan at all.
There is one major problem with Bookscan, which is that a fairly full access for one user ID costs a small fortune. Several thousands of dollars, in fact, so the cost is really prohibitive for a lot of literary agencies. There are some ways to get some information for less; but if you want to really have fun with this you've got to pay big bucks. For me, for that kind of thing, cost is almost no object because I've always been fascinated with numbers. In college, before I had a career in publishing, I would always enjoy looking at the index cards used to track ordering activity at the Community Newscenter locations in Ann Arbor, and it's not like I stopped doing that sort of thing once I actually had a real professional reason to do so. Imagine how happy I am every Wednesday morning when I can have a week of actual POS sales data for the entire JABberwocky list delivered to my computer.
And yes, the information is flawed. It doesn't include Wal-Mart, which guards its sales data zealously. It doesn't include Larry Smith's table at a convention. It doesn't include a lot of grocery or drug store or similar channels, though they've kind of gone from 0 to 25 MPH in the past couple years first by adding a few supermarket chains like Kroger and Stop and Shop and this January adding Hudson News. It doesn't include a lot of non-traditional channels, so I was out of luck the same way as Andrew Zack when I tried to market a sequel to a book that was selling mostly through Motherhood Maternity. Maybe if you as an agent subscribe to Bookscan you can be aware of possibe situations like this and try to address them with a line in your marketing letter, but that may or may not work. It could just force a house that wants to reject to come up with some other "polite" but silly reason to say "no." Or maybe they'll believe Bookscan before they believe the agent. Agents do lie sometimes. A baseball free agency season hardly goes by when some GM isn't complaining about the mysterious other suitors Scott Boras claims to have, and I'm sure some editors feel that way about some literary agents.
But Bookscan is a tool, and like all tools it can be used right and used wrong. Right about the time I first got my Bookscan subscription, I noticed that the just-published Crossover by Joel Shepherd was selling pretty danged well for a book that had been taken by Borders but not by Barnes & Noble. I was able to go to Lou Anders at Pyr and point this out, and Lou was able to go to the sales people at Prometheus Books, and they were able to go to B&N and get them to take some copies of the book. And this helped turn CROSSOVER from a dubious proposition into a solid enough performer that it's now been chosen to launch the mass market program from Pyr. A lot of things had to go right for this to happen, but it all started with somebody paying attention to those Bookscan numbers.
And now this year, I was able to notice a sales spike for Adam-Troy Castro's Emissaries from the Dead after it won the Philip K. Dick Award, and I was able to make a good numbers-based case to Diana Gill at Eos and she was able to go to her sales people, and we've gotten some renewed support from Borders for Emissaries and its sequel The Third Claw of God. It's too soon to know if this will have the long-term payoff that I saw with Joel Shepherd, but again this is at least getting some good use out of the tool.
Of course it costs so darned much, that I'm not sure on strict cost-benefit terms that I can justify how much I pay in order to achieve these victories. It's a no-brainer for me because of the weird wiring of the Joshua Bilmes brain. But does it really pull my fat out of the fire near often enough to start saying cost should be no object?