Data... It’s big, confusing, and hard to maneuver. Yet, the collection of big data has become an increasingly valuable resource for prioritizing work and making quick, effective decisions.
Your advantage comes when you know how to decipher this “big data.” Or, use tools to enable you to understand it better. For instance, you know your titles sell on Amazon because you get a report that tells you so. That’s big data. However, the report doesn’t really tell you the details or what to do about what you’re seeing – it is only descriptive.
Deciphering the data is when you can see data in context. For example:
You might be thinking, “Yeah, big deal. Those things would be nice to know but that would take hours of siphoning through Excel sheets.” You know what… you’re exactly right! But we did that for you.
Recently, our Consumer Insights team took 150,000 titles and over 100 million data points from our Marketing Insights software to decipher the data and see what stories it would tell. As Marketing Insights collects more than 75 data points per title every day and it tracks hundreds of thousands of titles for publishers, our deep dive took hours, thousands of rows in Excel, massive pivot tables…you get the picture. But, we felt the results would be worth knowing. And, in fact, the results are fascinating and actionable.
*Totals may be greater than 100% because some titles fit in multiple categories.
Better Understand Amazon
Better Understand Consumers
Understanding Publisher Behavior
Just as family members come from different directions to get to a reunion, the same is true for titles reaching higher sales volume. The destination is the same yet the path to get there is unique.
Marketing Insights is the GPS navigating you to your destination. We created the tool to translate book data into book sales by collecting large sums of information from Amazon, of course. But also relevant data from Goodreads, Google, Facebook, Wikipedia, and other key online platforms used by consumers.
This data is then deciphered and given to you through bite-sized, actionable marketing tactics for every title on your list.
Your lists are unique, skewing toward one format or another, so some of these insights will be more useful to you than others. Today, we’ll lay out a broad understanding of the Amazon Marketplace and Marketing Insights to show a vast host of you (publishers), no matter your genre or business model, can win with this information.
If you’d like to see how Marketing insights can guide you to sales growth, schedule a demo with our team.
Now, let’s dig into the details.
An oft-misunderstood and often high-value data point from Marketing Insights centers on the Amazon Buy Box, specifically the “lost” Buy Box. A lost Buy Box scenario is when a book is being sold or offered for rental on Amazon, but the actual transaction occurs with a third-party seller (someone other than Amazon).
This can impact publishers Amazon sell-through, account and channel strategy, as well as advertising, author relations, and much more.
In the example below, “CozyShades” would be the third-party seller, and this would be considered a lost Buy Box by Marketing Insights.
According to the Marketing Insights data below, from 2019 through 2021 (early pandemic fulfillment issues), there was a considerable spike in third-party sellers winning Buy Boxes on Amazon.
45% of the titles that Marketing Insights observed as being available for purchase on Amazon had a Buy Box featuring a third-party seller, not Amazon.
What we see here:
Looking at new, used, and rental, in the US, most of the third party Buy Boxes are going to Book Depository who is owned by Amazon but orders and sells books separately.
3rd Party Sellers (US)
For the UK, there is a similar story, though fewer sellers overall than in the US. In the new table, the top 10 sellers represented about 60% of the lost Buy Box cases. The story is the same for the sellers of used books too. The top ten sellers make up 60% of the market.
3rd Party Sellers (UK)
This is Amazon’s inventory messaging in the Buy Box. This is what the consumer sees. Here you see the top five messages we've seen Amazon say to consumers about the availability of a title.
These are the five most common availability messages you’ll see in the Buy Box.
This messaging can have a major impact on your book sales. From our data set, we looked at when consumers are seeing different availability messages, does that have any impact on the consumer's purchasing decision, the likelihood for that book to sell through, and the overall sales rank for the titles that have those messages?
We saw there's minimal effect on sales or sales rank when the availability message is “In stock” or “In stock soon.” No surprise.
There are some messages here that clearly do have a negative impact on consumers ordering the title. One of those is for pre-order titles.
“This title has not yet been released” as opposed to “This title will be released on.” The prior messaging will decrease your sales rank. Ensure you have a book’s release date present in your metadata and tell consumers it's available for pre-order. This will help improve your pre-order sales.
One myth we’re interested in busting concerns those “Only X left in stock” messages. We’ve often gotten the question: does Amazon message this way to trick consumers into feeling a sense of urgency?
With the often-implied assumption that “yes they do” and “hey, it might increase their likelihood for purchase.”
This is false. This messaging decreases sales rank by 183%. In the first chart, you can see Amazon has begun to decrease its use of the “X left in stock” messaging dramatically and replace that messaging with “In stock” even though the inventory is low.
The data strongly suggests that consumers are not more likely to buy a given book if they are told that Amazon might run out.
What’s the relationship between sales rank and units sold? This is a hot topic.
We mapped a book sale with the Amazon sales rank for the following day. If a title sold 10 units on Tuesday, we looked at how that correlated with the Amazon sales rank on the next day.
Units sold and Amazon sales rank have a clear connection.
There are other factors too. For instance, the sales history of the title and its average rate of movement over a period of time. Books with historically long sales histories can have lower sales for a period and maintain a good sales rank. If lower sales persist, then the book’s rank will start to bottom out.
In general, there's a strong relationship here. You can see the curve between Amazon sales rank and units sold is very steep.
On the print side, to achieve a rank of over 100, we were seeing titles selling in the hundreds of units per day. On the eBooks side, that's even sharper. Titles selling over 1000 units every day were those seeing ranking in the 100s in the Kindle Store. That's dominated by major bestselling titles and titles with current price promotions running (e.g. Kindle Daily Deals, etc.).
How many unit sales does it take to rank?
Sales Rank: Observations & Potential Actions
On Amazon your titles fit into categories based on the BISAC, BIC, or Thema codes in your metadata, plus off-page keywords, combined with user behavior on Amazon. Amazon will, essentially, re-shelf books if the demand seems to be based on the book “about” something other than what the publishers has said it is about. That said, publishers can exert a fair amount of control here and ranking in category is important for visibility to the right potential buyers.
Some categories are easier to rank in than others. This is a key insight to keep in mind. It’s easier to rank for many adult nonfiction categories, especially with low daily sell-through. Even more true is the more niche you get into each of the subcategories.
Unsurprisingly, genre fiction is competitive. The competition both on the eBook and the Print side for placements within these categories is steep.
We looked to what degree Amazon offered a discount to the consumer for both hardcover and paperback books across 2019, 2020, and 2021.
No discount is by far the most common scenario we encountered. Over 50% of all books we tracked were not discounted by Amazon to consumers. The shift begins as we look deeper into the discount status.
In 2020, you can see Amazon start to reduce their deep discounting, both on hardcovers and on paperbacks. There was more discounting at the 10% and 20% level, and a lot less discounting happening at 40% and 50% off.
In 2021, deep discounting began to increase again across more books. This is a positive sign for title sales in Amazon and, perhaps, a return to pre-pandemic discounting and an Amazon environment publishers are used to.
Knowing when Amazon is discounting is key because it removes friction from the consumer purchase. They're discounting to consumers and helping you with your sell-through.
We set out to find clustering and patterns in Kindle discounting and how Amazon is pricing eBooks to the consumer based on the publisher's list price. We refer to these discount clusters as “price bands.”
Remember, over 45% of the eBooks we track are not discounted at all. We did not isolate those being sold under the agency model (in which the publisher sets both the list and consumer price). In our data set the majority were being sold under the wholesale model, in which Amazon does set the ultimate consumer price. We do see some pretty clear patterns in the way that Amazon discounts eBooks.
The larger the bubble. The more often the discount occurred. We saw consistent discounts of 20% off, 25% off, and some at 40% off. We also saw patterns in what we’re calling “Horizontal Amazon Price Bands.”
Regardless of where you have set your price, our data suggests there are certain price points Amazon likes to provide to consumers. Above, you’re looking at titles priced up to $100 on the list side. And we saw a $23.74 price point jump out at us.
Now, it is not terribly common for $100 books to be priced down to $23.74 but we do see that happening. We repeatedly see Amazon moving toward these specific price points, likely because Amazon shoppers are attracted to them, and they help increase sales.
The other thing we wanted to look at was the publisher price bands. These are the price points publishers like you are typically pricing your books along with the typical Amazon discount. In the $100 price range, we see a lot happening at $59.99 and $39.99. That's typically more on the professional and academic content. On the trade side we start to see more at the $19.99 and $24.99 price points.
The most common publisher price band we saw under $20 was $9.99. Over 10% of the titles that we looked at in this price band are priced at $9.99. Amazon frequently discounts eBooks priced at $9.99. We see them bring the price down into the $7 and $8 range. The average discount here was about 13%.
Also, we saw that eBooks priced by publishers at $7.99 are likely to stay at $7.99. Only 7% of eBooks were discounted for an average of $.48.
Understanding how Amazon may discount your eBooks is crucial. When you have control over something like the list price and knowledge of Amazon discount practices you can make more strategic decisions about what the best price for profitability is for you for a given title.
Amazon doesn’t have formal price bands, but this is what we found in the data.
Kindle Prices: Observations & Potential Actions
Seven and eight we combined. We know... we cheated. Combining Amazon and Goodreads consumer reviews and ratings is more informative than trying to look at them separately.
Each of these charts is the distribution of star ratings across Amazon and Goodreads and how that relates to overall Amazon bestseller rank.
We set out to understand the degree consumer review and rating behavior are different in these two environments and their relationship to how much a book might sell. Overall, books on Amazon are more highly rated than those on Goodreads. Intuitively not terribly surprising; Amazon is a commercial environment and Goodreads more of a non-transactional “reading” environment.
High selling titles or bestsellers with an Amazon rank between 1 and 1,000 have a tighter distribution of star ratings. A star rating of five doesn’t necessarily make a big difference.
It doesn't hurt. But what is more correlated with sales success for a book is many reviews or a consistent and a growing number of reviews that maintain a high star rating.
You won’t see a lot of top selling titles with one star. It doesn't happen. In fact, very few top selling books have fewer than three stars ever on Amazon. That happens more on Goodreads.
So, what can you do about this?
What do we mean by “product page?”
Notice we circled “Look inside” and the description. You can see a bolded headline and paragraph breaks between the next two lines. These are important items on the product page.
We look a lot at this optimization because we've seen a ton of studies that talk about how much it matters to a consumer converting to sale when they reach a page.
Grade 1 = bad
Grade 2 = good
Grade 3 = great
Product Detail Page: Elements that matter
We wanted to ask the question of ourselves. Does this make a difference? Does it relate or have any relationship with how well a book might sell on Amazon?
Titles having a Grade 3 also rank better. The top three elements correlated with higher sales rank being:
We’ve observed this in smaller controlled studies as well that hint much more at causality, but this data suggests that you try to make the description headline bold and 10-12 words in length.
You don't want a huge paragraph, just something short and punchy to grab a reader's attention before they click through to read more. Eye and mouse-mapping studies support that.
Make sure the book description has paragraph breaks. Format so readers can parse it. Each sentence leads to the next so consumers want to read more about your book.
This was certainly gratifying for us and validating in terms of some of the things we have worked on and suggest publishers do through Marketing Insights.
Book marketers are working on their metadata all the time and guess what — it makes a qualitative and quantitative difference. Better consumer experience and better sales do go together.
Product Page: Observations & Potential Actions