Thursday, February 15, 2007

Letter From the ECPA


Recently I received this letter from Michael Covington, Information and Education Director for ECPA (Evangelical Christian Publishers Association). Michael very kindly took the time to point out some erroneous statements in my previous posts about data reporting for bestseller lists. (Links to these posts in the sidebar.) He gave me permission to run his letter publicly to clear up these misconceptions. I thanked him privately, but here’s another, public thanks from me for his time and energy. It’s hard researching these issues from the outside and getting everything right, so I’m very happy to have an “insider” tell me—and now all of you—what’s really up.
---------------------------
Dear Mrs. Collins:

… Through reading your blog, I can sense that you are an influencer in the world of Christian content, specifically in the fiction genre. I have also noticed that you have a keen awareness of and interest in book sales data and bestseller lists. In my current role I am involved in the development of Pubtrack and all data initiatives for ECPA members. Because of my roles in Christian retail and the world of book sales data I thought it might be helpful for you to have some insider information that would further educate your blogging endeavors.

In a couple of your blog posts you commented that STATS was the industry’s first data collection service, then CROSS:SCAN was launched by CBA and then Pubtrack was a “back atcha” to CROSS:SCAN. I was actually on the CBA B.O.D. when CROSS:SCAN was launched and can tell you that while it may have appeared that Pubtrack was an answer to CROSS:SCAN, it actually had been in the works for quite some time.

STATS was developed in 1999 to aggregate book and Bible sales data from Christian retailers in order to help Christian publishers and retailers more fully understand how books were performing out the front door of Christian stores vs. simply into the back door (as is reported by AAP). In 2004 STATS Plus was launched with increased search and analysis functionality and then Pubtrack was first announced in January of 2006.

As you have rightly noted, STATS at one time did have over 900 rooftops reporting, which included Family Christian Stores. As a practice, we have never announced who is “in” or “out” of STATS data, however in this case the retailer made the announcement that they were withdrawing. In your post dated
May 4, 2006 you wrote:

“Remember that the sales above were only reported from about 1000+ Christian bookstores. For a rough estimate of total sales of the book in a given month, the formula I’ve heard is to multiply the STATS number by 3. Of course, from current lists this formula would be a higher multiple, since the number of stores reporting to STATS has fallen from 1000+ down to around 650.”

The actual multiplier we have used has been 2.45 for books and 1.45 for Bibles, however contrary to what you wrote; the multiplier never changes for STATS (or for Pubtrack), regardless of the actual number of stores reporting. This is because a normalization factor is included in the output data in order to provide comparative sales analysis.

So, when you look at STATS data you are looking at a virtual 1000 Christian retail stores. This method has been proven accurate +/- 4% and it helps those evaluating the data to look at numbers week over week or month over month to compare apples to apples. Otherwise if you have 532 stores send in data one week and then 674 the next week, you cannot compare aggregate numbers to one another for trends analysis. Therefore, we insert a normalization factor into the tables that averages out the numbers so that you are always looking at “1000 stores,” therefore the multiplier to figure total channel volume never changes. As I understand it, Bookscan uses a similar normalizing number in their data for the general trade.

Going forward, our data collection tool as you have noted, is called Pubtrack. Last week marked the official rollout of the Pubtrack user interface and publishers and retailers will begin using the tool as early as next week. Since this is similar to a beta-phase, we will keep STATS up and running through the end of this year. You commented multiple times that it appeared Pubtrack never materialized. I can understand how you came to that conclusion; however Pubtrack has been around in a tangible manner for more than eight months and has undergone much iteration to bring it to where it is today.

A logical question might be “Why has it taken so long to come to some form of arrangement between CBA and ECPA?” As you have pointed out in some of your more recent posts, the two associations have continued to work together in this area of data since the announcement of CROSS:SCAN. In one post, you mentioned a letter from CBA that referenced, “more news to report in January during Advance 2007!” While we were hopeful that something definitive could have been announced last week, I am not sure that Mr. Johnson was making a direct reference to the CBA/ECPA talks as CBA has made another
announcement about an agreement between CROSS:SCAN and industry newcomer Above the Treeline.

Finally (I bet you hoped I would get there), with regards to your comments on bestseller lists, ECPA is keenly interested in tracking books wherever books may be sold and have as one of our strategic goals to implement a multi-channel bestseller list by compiling sales data from all channels. However, as you can see navigating the relational waters with all of the different entities involved is a Herculean task and one that we anticipate could take years to complete.

Ms. Collins, thank you for your ministry and gifting, we appreciate the voice you have in this industry and pray that we can be a source of information for you in the future.

Warmest regards,

Michael Covington
-------------------------

Now wasn't this guy tactful and kind, even while pointing out my mistakes? I sure thought so. Very impressive. Again, I send Michael Covington many thanks for this information.

7 comments:

Tina Ann Forkner said...

What a nice guy. And great information! You are super for posting it, Brandilyn.

Karen said...

Am I the only one who didn't get all of his publishing mumbojumbo? This reminds me why I didn't major in math. Numbers are evil, well sort of (actually not really). But I definitely salute the people who enjoy working with them. I thank God for them!! He did correct you in a very nice way. So he gets major cool points for that.

But the part that got my attention was: "Through reading your blog, I can sense that you are an influencer in the world of Christian content, specifically in the fiction genre." Yay for you! And yes, your blog totally rocks, "Mrs. Influential". ;)

If you could break down what he said into laymens terms I'll give you two chocolate covered thumbs up!

Eden said...

He's a nice gent to help us all understand everything from the insiders view. Thank you so much for posting this for us all.

~ Brandilyn Collins said...

Karen, I have to admit my eyes kinda crossed at the "normalization factor" stuff. I had to read through a couple of times to get it.

As I understand it (sheesh, I hope I don't make ANOTHER mistake), no matter how many stores report to STATS, the data are extrapolated as if 1000 stores reported. This really makes no difference to anyone reading just the bestseller list. It makes a difference when you're an author reading actual STATS data on the number of units sold of your book that month. Since all stores don't report, obviously all sales of a title aren't showing up--so you have to multiply the STATS number for a clearer picture of how many units are really selling off shelves. Michael C. is saying that BEFORE they report the data, they "normalize" it to represent what 1000 stores would report. So the multiplying factor for actual sales of a book never changes. It's always 2.45. And he says this method has proven accurate by plus or minus 4%.

That's really interesting info--and nothing that an editor or agent had ever told me. That's why I so appreciated MC's letter. Only someone with full knowledge of how this stuff works could know this.

As for the "influencer" sentence, to be honest, I actually chalked that up to his being tactful and kind--more than accurate. :)

Kristy Dykes said...

A lot of technical stuff. Thanks, B., for sharing it with us.

Cara Putman said...

What great information. Glad he wrote and then allowed you to share that with us, Brandilyn.

Karen said...

Brandilyn, thanks for "summing it up". Now, I get what he's saying.