Bookshop.org doesn't work the way you think
So It Turns Out, #43
Hi,
So here are some things I’m thinking about this week.
So It Turns Out Bookshop.org Doesn’t Work the Way You Think
In general, I think it’s probably best for this newsletter when, after having had the same conversation a few times in a row explaining something cool and new, I just write down that conversation and send it to all of you. This is one of those.
You may be familiar with bookshop.org, the sort-of friendly alternative to Amazon.com (at least, for purchasing books). At the beginning of the pandemic, BookShop (which coincidentally, launched in January 2020), became popular among folks who were trying to ensure that pandemic lockdown shopping didn’t only benefit large companies, and kept smaller companies afloat. This meant that BookShop grew dramatically in the space of just a few months, and had a reported $50M of revenue in 2020.
Before I looked into it, I assumed that BookShop works like, well, Amazon did in its earliest days: BookShop takes an order from Alice, who passes it to Bob’s Independent Local Bookstore (Bob’s ILB). Bob’s ILB fulfills the order from the books it has on its shelves in its shuttered store, mailing it directly to you, the customer. BookShop then takes a cut off the top for being the marketplace that matched buyer and seller. After all, it exists to keep local bookstores alive, local bookstores have products on their shelves that they can’t sell, that must be it, right?
That is, as they say, completely compelling, and totally and utterly wrong, as I recently learned when I tried to use a bookshop.org gift card in person at my local independent bookstore.
Instead, BookShop has adopted a model that is substantially more interesting, and unlocks some really interesting opportunities for the Independent Local Bookstores it works with.
We all understand (hopefully) that much of the Web is funded on the back of advertising. What many people don’t necessarily understand, is that (oversimplifying dramatically), there are two types of advertising, usually called “brand” and “performance” advertising.
Brand advertising is, well, what Coca-Cola mainly does — reminding you that Coca-Cola should be associated with themes of coolness, refreshment, and polar bears, to make you prefer Coca-Cola’s brand when you’re making purchasing decisions generally.1 This is usually paid for on the basis of “impressions,” which just means how many people see the ad, with some scaling factor depending on how valuable each one of those eyeballs is to the ad space purchaser. For example, an ad running next to content viewed by folks with low purchasing power is less valuable than an ad running next to, e.g., a golf or tennis tournament, which tends to be watched by folks with lots of disposable income.
Performance advertising, by contrast, is advertising intended to drive an immediate action by the person seeing the advertising. A classic example is the sort of ads you see on Instagram, which are intended to get you to go to a website outside of Instagram (or increasingly, inside of it), to Buy Now sunglasses or a shirt or skincare or something. These can be valued on a traditional impressions basis, or, increasingly, a pay-for-success basis (e.g., $X per person who sees the ad and installs the Uber Eats app on their phone)
While a full discussion of this industry is well beyond the scope of this newsletter (LUMA Partners’ LUMAScape is considered canonical for listing the major players in the space, and even most ad executives don’t know what most companies in there do), let’s focus in on one small aspect of performance advertising: affiliate marketing.
You may already be familiar with affiliate marketing, even if you don’t know that term. Affiliate marketing is performance advertising being done on a pay-for-success basis by a given known advertiser. That might be opaque, but you know plenty of examples of this. The purest visible example is probably the NYT’s Wirecutter2 product review site, where the site reviews a category of products (e.g., printers), and offers links to those recommended models. Each link, buried deep in its text, contains a referral code (for example, “?tag=thewire06-20”) that tells the selling website “oh, this user came from Wirecutter,” and allows it to attribute any business coming from that link to Wirecutter, and pay it a performance fee (usually ~5% of all sales) attributable to that link.3
This is also quite common in the blogging space; as a reminder disclosure, this Substack uses Amazon affiliate links whenever posting links to books or similar things you can buy on Amazon. And Amazon not only pays out for the specific book I recommended; if you go to Amazon and also pick up some AA batteries in the same trip, I might get affiliate fees on them as well.4 As disclosed in the footer of all issues of this Stack where I’ve used those, I donate all (quite immaterial) profits from them to charity. (Additional disclosure: I’ll probably start playing around with Bookshop affiliate links as well)
You even see it in the YouTube and Twitch streaming spaces, where a given video creator might tell you to use their referral code or link when purchasing from a given website.5
Okay, so affiliate marketing allows content producers to monetize referrals of their readers/viewers to sites where they can buy things. And everyday, across the web, small slices of ~5% of transactions are seamlessly paid out to those who refer products to others, and Amazon and its compatriots only pay when it works.
Your natural question, then, might be: why is a competitor to Amazon in this space even possible? Why would any referrer switch from the Amazon referral ecosystem, which presumably can afford to pay a bunch and have lots of products worth referring, to this new start-up?
And that’s where the genius of Bookshop’s approach comes in. They, essentially, looked at Amazon’s vast profits on books, and said, “hey, we’re trying to become a Certified B Corp, what if we didn’t care as much about profits as being a utility for independent bookstores that covers its costs and makes a little profit? Just how big of a affiliate referral fee could we afford to pay?”
And the answer is: a ludicrous amount. Because BookShop isn’t taking inventory risk, and neither is the independent bookstore; rather, a major book distributor, Ingram, is the one buying, warehousing, packaging, and shipping every order. BookShop’s just the front end.
As bookshop.org describes it, there are three different fee structures6:
If you buy a book from a specific, bookstore-branded BookShop link, that particular bookstore gets 30% of the cover price of the book; for the remaining 70%, about 10% is lost in discounting and promotional pricing, about 10% goes to cover payment processing and fulfillment, and about 50% goes to Ingram, who actually physically holds and ships the copy of the book. (Ingram, of course, in turn pays a large portion of that price to the publisher, who pays a portion of that fee in turn to the author, who in turn pays a small portion to their agent, etc., etc.) Crucially, BookShop says it turns no profit on this performance ad category — to the extent it benefits BookShop instead of bookstores, it’s purely brand advertising to make you feel good about BookShop helping booksellers.
If you buy a book through any other affiliate link (for example, if I had one), the initial 30% chunk is instead split this way: the affiliate gets 10%, 10% goes into a profit pool split across independent bookstores7, 10% goes to Bookshop for profit. Then, it seems the other 70% is split as above. This performance advertising channel drives the activity where BookShop says it generates the minority of its revenues, but the majority of its profits.
Finally, it appears that if you just go straight to bookshop.org and buy a book, the math is the same as “any other affiliate link”, but BookShop captures the 10% affiliate fee for itself. This one seems fuzzier, in their telling, about how big it is, and certainly I’m sure they hope to grow it over time.
So, is this a sweet deal for bookstores? After all, the math above also implies that, outside of a BookShop transaction, a bookstore normally is making about 40-50% gross margin on each physical book they sell from their shelves. Here, they’re making only 30% from these BookShop guys. Why would they want to do this?
Well, here’s the thing: they’re making money off of books they never have to buy.
Now, if you’re someone who’s been a consultant, banker, or has been employed by or owned an inventory-heavy business, you’re probably now shouting at the screen “NO WORKING CAPITAL? WHAT THAT’S AMAZING.” If not, you’re probably totally confused why that’s awesome. Let me explain.
The average bookstore has to, well, have books in it. Therefore, you need a chunk of change to buy books wholesale to sell retail. And once you sell a book, a huge chunk of the revenue has to go back into the business to buy more books to sell.8 The faster you sell each book, the more times your inventory turns over a year, the better your profits are; if you don’t turn over your inventory fast, you have money sitting useless on your shelves (or perhaps even depreciating, if that book is hot now, but old news in a year’s time).
But what if for some of the books you sold, you never needed to have them on your shelves at all? What if you never needed to buy them at all, because Ingram owned them and bookshop.org sold them, and you just, basically got a fee for the good vibes your brand brings to bookshop.org? That unlocks, effectively, unlimited upside.
Of course, yes this is a lower overall margin per sale that could compete with you, so if you were really really good at managing working capital, you might not be thrilled. There are some bookstores that feel this way. But most bookstores don’t get into the business of selling books in hopes of managing working capital models and inventory turns. They’re in the business of helping people find books that bring them joy.
Similarly, you might worry that bookshop.org might one day pull a Darth Vader and announce, “I have altered the deal to Amazon-like referral fees, pray I do not alter it any further.” And then you’ve helped build up a competitor to your business, with scary internet scale. Again, some bookstores do feel this way, and it’s no doubt a concern, but since Amazon already exists, and bookstores compete right now with Amazon by buying books from distributers like Ingram anyways, this is less of a concern for many booksellers than at first glance.
So where might BookShop take us? Honestly, I don’t know. But here’s one idea I can’t shake. BookShop is, essentially, paying people to get really, really good at marketing their independent bookstore’s vibes. If you knew you were getting paid 30% of every sale, on inventory you didn’t have to own, just what could you do to make that bookstore famously awesome, like a Strand or a Shakespeare & Co for the digital age? What events could you hold? What physical and digital presence would you want? How would you change your inventory mix on your shelves, if you were trying to Build A Memorable Brand slightly more than than trying to sell a physical book?
What, in other words, are the boundaries of the bookstore model when you have zero-marginal-cost upside opportunities?
I don’t know, but I’m curious to find out. (And, disclosure, I’ll probably play with their affiliate model some, myself.)
May this bird bless your timeline
May these manatees eat Skyline Chili
Your moment of joy
Disclosures:
Views are my own and do not represent those of current or former clients, employers, friends, or my cat.
I may on occasion use Amazon Affiliate or similar links when referencing things I’d tell you about anyways. As an Amazon Associate I earn from qualifying purchases; I donate the proceeds to charity. While Substack has a paid subscription option, I don’t have any plans to use it at this time and anyone who gets this newsletter now surely won’t be ever paying for their subscription.
Obviously, given the logic I just laid out above, I’m going to consider being a BookShop affiliate, too.
With the possible counterexample of the Coca-Cola ads you see in a movie theatre before the trailers, which probably are in part intended to make you buy a coke right now. But again, that’s also probably to make you associate Coca-Cola with the feelings of escape, relaxation, etc. you get at the movies.
This stuff gets tricky; contrary to popular belief, the ad industry is able to recruit PhDs not only because Something Is Broken With Academia, but also because there are actually really fun causal inference problems in it, too.
(That being said, Something Is Broken With Academia.)
With congratulations to their staff on their successful unionization.
Yes, if you follow the online ad space, you are cringing right now at how blithely I’m implying you can just attribute business to a given ad campaign without lots of challenges at scale. But for purposes of this discussion, I’m intentionally burying that under the hood.
Amazon doesn’t tell me who bought a given product, incidentally. That would be creepy.
For these, I understand there is sometimes a combo of base ad fee + performance-based affiliate fees for the sales driven to the business, which you might think of as a combo of brand and performance marketing goals.
They describe it as two, but it’s cleaner to break out this way, I think.
The specific formula for divvying up funds is unclear (e.g., pro-rata, weighted by sales, etc.). BookShop seems to indicate that a bookstore can benefit from this pool even if they don’t sell books on BookShop, but it’s unclear exactly how that works.
Yes, this analysis gets slightly better if the local bookstore is special-ordering a book for you online, and yes, I’m grossly oversimplifying a range of concerns about Net X days payment terms, lines of credit, etc., but the core problem — a bookstore needs to spend money to make money — is a real one, and BookShop has an interesting solve here.