Pub Rants

Bookscan Is Great—Except When It’s Not

 16 Comments |  Share This:    

STATUS: I’m sure I don’t have to say that yesterday was a little hectic.

What’s playing on the XM or iPod right now? SIMPLE DAY by One Eskimo

As an agency, we have a subscription to Bookscan and every Wednesday, we send out a sales reports in Excel spreadsheets to each of our clients for their published titles. (FYI–Nielsen is a subscription service that captures point-of-sales information from certain retail outlets.)

This sounds great. Real sales numbers! Except not every retailer reports to Bookscan. Some key accounts like Costco do report but other key accounts like Walmart do not.

Which means that Bookscan is not a whole picture of how a title is doing.

So over the years, I’ve created our own system of calculating how accurate it is by comparing the royalty statement sales to the Bookscan number sales and capturing the percentage difference.

For some genres, it can be off by 50 or 60%. That’s a lot. The numbers for literary fiction tend to be a bit more on target as Bookscan seems to capture about 70% of sales for this segment.

Why is this important? Well, if you are a midlist author looking to move houses, well, guess what numbers the editors are looking at in order to base a decision of whether they want to offer for you or not? You guessed it. Bookscan.

And if that number is only capturing 50% of the sales… I have to firmly argue the actual sales numbers and sometimes, that doesn’t matter. The house will often make a decision based solely on those Bookscan numbers. Hugely frustrating as you can imagine.

By the way, Bookscan does not currently capture digital point-of-sales. Yeah, that’s going to need to change as more and more sales are done digitally in the upcoming years. And yet another problem with Publishers deciding that Bookscan is a reliable reflection of sales…


16 Responses

  1. Anne R. Allen said:

    Thanks so much for making this clear. I never could figure out why Bookscan is so off on some sales. It does seem ridiculously unfair, especially for romance and mystery writers who sell in non-bookstore venues.

  2. zz said:

    The dream of publishing sounds glamorous, but the business of publishing sounds PAINFUL. Can’t wait 🙂

  3. James Montgomery Jackson said:

    Sometimes as outsiders we shake our heads at the way a publishing decision is made.

    And yet, if a publisher compares Author A’s Bookscan information to Author B’s (in the same genre) then the publisher can make reasonable decisions based on admittedly incomplete information.

    If I were a publisher the bigger issue I would be trying to figure out is how to capture e-book market information — and to project what that will mean to the author’s ability to generate revenue in times of flux.

    ~ Jim

  4. Kristin Laughtin said:

    By digital point-of-sales, am I correct in assuming you mean purchasing books off Amazon and other sites?

    Yeah, that’s a huuuuuge market. Walmart is as well. Bookscan definitely needs revamping, or publishers need to wake up on this issue.

  5. Jill said:

    So Kristen, what’s the best way to buy a book, if one of my goals is to support my favorite new authors (and by extension, their caring, blogging agents)?

  6. Anthony said:

    Using flawed statistical models to formulate business decisions in the face of evidence the model is wrong is indicative of greater problems below the surface.

    I would assert even the most respectable publisher engaged in this monkey business (and it is just that: monkey business) is either currently going to face a business crisis or will in the future face a business crisis.

  7. Eric Riback said:

    OK, I realize a publisher’s sell-in is different than sell-through. But they have to know what their NET sales are for a given title and what the Bookscan reports say over the life of a hardcover. How can it be that the publishing houses aren’t interested in or capable of the same kind of analysis your li’l ol’ agency is doing?

  8. Dustin said:

    Great post.
    I work in the video game industry and we are haunted by a similar process. In the gamez biz it’s called TRST data, and the same flaws exist.

    One of the interesting differences is that we NEVER get the multi-title deals and new projects are evaluated and budgeted on early TRST data trends. This can really sting because these (incomplete) numbers are verifiable (which publishing and distribution love) and our projections are not. If you get a slow first three months out of the gate then your budget to build a sequel is compromised. It becomes a self-fulfilling prophecy that can kill a franchise overnight.

    Kristin, do publishers cut or drop funding for a title within a series during the run based on poor Bookscan results? I can see that being equally frustrating and completely believable.

    After following your blog for a year now, I’ve come to realize that the process of selecting, producing, funding, evaluating, adjusting, marketing, developing, publishing and distributing books seem to mirror that games industry very closely.

    I tip my hat. Misery and company and all that fancy jazz.

  9. Anonymous said:

    Of course publishers are aware of this – Bookscan’s a sample, everyone knows that. They ‘only’ capture 70% of sales? Well … TV ratings for three hundred million Americans are extrapolated from a few thousand homes, and advertisers and program makers make hundred-million-dollar decisions based on those.

    There are going to be some types of book that do a little better, and some that do a little worse under Bookscan. There are some categories – like Amish romance, say, or End Times Christian prophecy – that pack the supermarket shelves but barely flicker on Bookscan.

    It’s not a perfect system … well, OK, show me the perfect system. The one that takes into account the people from the US who bought the third ‘The Girl Who …’ book online from the UK because it came out five months earlier there, or who bought a book from Goodwill, or direct from a small press.

    No one treats Bookscan as the exact figure of all copies sold. They use it as a measure of how *well* a book sold, compared with other titles.

  10. Red Boot Pearl said:

    Hmmm…that’s kind of depressing. But asi es la vida, There will always be something else to worry about when actually published.
    It’s kind of funny how many people think getting published will solve all their problems. Between deadlines, the curse of the second novel, worrying about people showing up for signings, and the constant nail biting over sales people go through, I’m really starting to cherish this “pre-published” phase.

  11. Anonymous said:

    You are unfortunately stuck with the standard used by pub houses. Frustrating to be sure but there it is. As for Bookscan, I would expect they will adapt to tallying sales that include those areas like digital point-of-sales they miss now. the industry is in a state of flux. Perhaps the greatest confluence of change since ol’ Gutenburg invented the press.

  12. behlerblog said:

    Ach, Bookscan – don’t get me started. We quit using them a couple years ago because they aren’t a realistic reflection of actual sales. So why bother? In truth, nothing can track every sale. Publishers should consider the work currently being submitted and their belief it’ll sell.

  13. Allison Brennan said:

    Bookscan is roughly 25-30% of my first 4 week sales. It does not include most groceries or many of the small drug store vendors as well as walmart, so mass market authors are really undercounted. When I did an analysis, bookscan was within 5% of target+bn+borders+bamm+airports+amazon.

  14. Annette Lyon said:

    Yep–the two chains that carry my books most don’t report to BookScan, and then there’s the Walmart thing. If you go by BookScan numbers, you’d think I sell maybe dozens or a few hundred books instead of the actual thousands I do.