Thursday, January 17, 2008

Understanding Reserves for Returns

I’ve been getting a few questions lately about reserve for returns and thought a blog entry would answer them best.

It’s an established fact that bookstores only sell about half of what publishers ship to them. Thus, when a publisher ships 100,000 copies, in all likelihood they will get 50,000 returns. Of course some books do better and some do worse, but as a back-of-the-envelope number 50% works.

Every publisher I’ve ever done business with establishes a “reserve for returns” for each book published. In some cases, this is a contractual number, but usually it is not. In some cases, the contract may call for its elimination after four or six periods, but often not (it really depends on whether your agent negotiates it). With nearly every publisher, an author can complain about how high the reserve for returns is and many publishers will adjust it if the book has been out for a while.

So let’s take a book that ships 20,000 copies and is published March 1st. And let’s assume the publishers royalty periods end 6/30 and 12/31, with statements issued 9/30 and 3/31.

When the first period ends 6/30, the publisher has four months of sales data. It shows 20,000 shipped and 2,000 returns. Thus, the book has a 90% sell-through. Now this is obviously way above average. The publisher knows that more returns will come. And, in fact, since it doesn’t have to report until 9/30, it has at least a couple of more months to watch the data and see what happens. What it sees is another 3,000 returns, say.

So the publisher now knows that in addition to the 2,000 returns in the first period, there will be at least 3,000 more in the second period. And that’s fine. The publisher knows the average sell-through is 50%. So far, it’s running 75%.

But publishers like to hedge their bets. Rather than anticipate just a 50% return, it may anticipate a 75% return. Better safe than sorry, right? So the publisher establishes a “reserve for returns” of an additional 15,000 copies as of the end of the first period (this would include the 3,000 it already knows were returned since the end of the first period and when the statement is going out) and reports a “net sale” of 3,000 copies, which is highly unlikely to earn out the advance unless it was fairly low.

Now the second period ends on 12/31 and the publisher looks at the numbers. Sure enough, returns continued to mount. They got a total of 6,000 in the second period. But, wait, it also got reorders! How great is that? It got 4,000 reorders in the second period. That means it shipped 4,000 more copies, so the net sale is a negative 2,000 copies for the second period alone. But more likely than not, the 6,000 returns were mostly part of the original 20,000 shipped.

The publisher can again look at the next period for a couple of months before it has to report on 3/31. It watches the returns and new shipments. Of the 24,000 total shipped, it has gotten back 11,000 copies and the book now has a net sell-through of 54%. That’s pretty good, but still a bit higher than average.

So the publisher looks at the reserve for returns and sees that 15,000 copies was a bit high. It looks at the second period and sees 4,000 reorders and 6,000 returns. It does the math and finds the book has a total sell-through of 16,000 copies to date or 66%. Still higher than average.

And then it looks at the two months since the period close and sees, hmm, 1,500 reorders and 1,000 returns. That’s only a 33% sell-through for those two months, but reorders are greater than returns. It releases or adjusts the 15,000 copy reserves so that at the end of the second period, it’s only holding a reserve of 7,500 copies, which it judges to be more than enough to cover any anticipated returns in the next period. Depending on the advance paid, this may result in additional royalty income to the author.

What’s going through the royalty manager’s mind is that of the original 20,000 shipped, they’ve probably gotten back most of the returns they are going to get. After all, the book has been on the shelves for a year. Most of the copies have sold or been returned. Now there are accounts that are reordering the book because they want to keep a copy or two in stock. The author lives in Chicago, so Barnes & Noble has “modeled” it to be sure there are two copies in each of their stores in the Chicago area (twenty-eight within fifty miles), since books are more likely to sell in an author’s home town than elsewhere. Sure, some of those copies might be returned, but a far lower percentage than of the original 20,000 shipped.

At the end of the third period, 6/30, again, the publisher sees 1,400 reorders and 600 returns. Sell-through for the period is 57%. Total sell through is 66%.

At the end of period four, 12/31, the book has been out twenty-one months. Reorders are 1,200 copies and returns are 300 copies, for a net sell-through of 75%. Total sell-through is still 66%.

Based on all of this information, the publisher feels safe lowering the reserve for returns. If I were this publisher and I saw a consistent 66% total sell-through and a 75% rate for the period, I’d probably hold back another 15% as a reserve for returns and pay the author royalties on 60% of the 1,200 copies shipped. In doing so, I’d be betting that most of the returns from the original 20,000 shipped have come back already, and that average sell-through on reorders will be about 66% for this title—higher than average!—as booksellers tend not to reorder books unless they have a good history with the book selling. At this point, the book is now a “backlist” title.

As a backlist title, the book can be expected to sell at a regular level every month. And returns either start to level out themselves or even slowly get smaller because only those bookstores that actually have a history of selling the book well are reordering.

In time, the reorders may eat up all stock of the book, and then the publisher has to decide if it is worth reprinting or to simply start fulfilling the occasional order by print-on-demand copies. Or even to just let it go out-of-print.

I represent a nonfiction book published in June 1998, nearly ten years ago!

For the six months ending April 30, 2007, it had the following sales:

869 gross of the $7.99 edition;

(2) gross of the 6.99 edition;

(22) returns of the $7.99 edition;

(39) returns of the $6.00 edition;

The publisher holds no reserve for returns of this title, because clearly sales are far in excess of returns. They aren’t concerned about a big return, as these copies are all reorders. And obviously they’ve reprinted, since there’s been a $6.99 edition and a $7.99 edition. The book has a lifetime sell-through of 73%.

Yet this book was no best-seller. It has made no one rich. But it did earn out its advance nearly twice. And it continues to sell close to 1,800 copies per year. I assure you that the publisher is perfectly happy with this title. Heck, looking at it, I’m thinking maybe it’s time to talk about an updated version!

In the end, reserves for returns that are adjusted regularly are a perfectly reasonable way for the publisher to avoid overpaying royalties. And authors should want publishers to avoid overpaying royalties, because it means that there’s more money to be spent acquiring new titles. Now and then a publisher screws up or gets a bit aggressive with the reserve and at that point the agent can call up and ask that it be released. In my years as an agent, I’ve gotten publishers to release tens of thousands in reserves. In some cases, this was fine. In at least two cases, subsequent returns demonstrated the publisher should have held onto at least some of those reserves. But publishers can still make a profit, even in those cases. More on that soon.