Over the last month numerous stories have broken suggesting that ebook subscription platforms might be facing a few roadblocks in their path to dominating the way that consumers buy and read electronically. Amazon's Kindle Unlimited announced that it would begin to pay authors by the page; Scribd began a major rationalisation of the number of romance titles in its catalogue; and a legal ruling in France required all ebook subscription platforms to drop the 'all you can eat' proposition that defines them elsewhere around the world.

Both Amazon and Scribd's decision to change what they stock in their catalogue and how they pay for it suggest two things. Firstly, it says that both services are beginning to achieve significant scale. Neither Scribd nor Amazon would be changing their terms of business if they were comparatively under-used. This chimes with previous estimations we've run on this blog that Kindle Unlimited at least drives millions of ebook downloads every month. The biggest challenge they face stems from the other inference we can draw from Amazon and Scribd's decision to change their terms of business. This is that they potentially have an issue with unprofitable customers.

One of the highly attractive qualities about subscription-based businesses to their owners and investors is that they provide regular income. The attraction to a consumer is different but related: especially if they offer all-you-can-eat access to content or a service as Scribd and KU do. Offering your customers the opportunity to consume as much as they can for a flat fee does involve an inherent risk if (as with KU and Scribd) you have suppliers to manage at the other end of your business. What happens when you charge $10 a month to your customers, but the majority of your customers consume $11 worth of books in that period?

At first it looked as though the ebook subscription service had this problem under control. Everyone assumed when they first launched that ebook subscription would look like the gym business. Here every customer pays the same flat monthly fee, but the gym owner makes the majority of their money from the people who pay their $50 a month to use the pool a couple of times, rather than the people who work out four or five times a week. The heavy users may be loss-making, but in the long-term this is more than compensated for by the light service users.

The signals that both Amazon and Scribd are sending to the market, however, are that ebook subscription services are less like gyms and more like all-you-can-eat buffets. Their decisions tell us that they have managed to attract a lot of unprofitable customers who are gorging themselves on the content in their catalogues, but rather fewer snackers who can be relied on to pay their $10 a month but not use the service regularly. And according to Scribd at least, romance readers are the heaviest consumers of all. Both therefore need to do something to manage either the supply of content to these heavy eaters (as Scribd has attempted) or the way it pays publishers and authors for the content that readers consume (as Amazon has done) in order to rebalance their services

Both decisions have drawn strong reactions from the community most affected by them: self-published authors. We know this for sure with Amazon, as its announcement about switching to a per-page payment for KU content was directed at authors with content enrolled in the programme via Kindle Direct Publishing. This platform is overwhelmingly used by self-published authors, and only authors who have chosen to make their ebooks available exclusively through Amazon are enrolled in Kindle Unlimited. The changes Amazon made to the way it paid authors were, in part, intended to address instances where self-published authors were gaming the system, releasing short-form content into KU because they received the same flat fee for a 'borrow' of a 30,000 word work that they would for a 130,000 word novel. Amazon's decision has been widely criticised as an instance of the company turning on the independent authors who have arguably been the main engine of ebook growth in the past year.

Similarly Scribd's decision to reduce the number of romance titles in its catalog has also primarily affected self-published authors. Indeed this story first broke when Mark Coker, the founder of ebook aggregator Smashwords wrote a blog post where he said that 80-90% of the (mainly self-published) romance titles that it supplies Scribd would be cut from the catalog. Smashwords was one of the first content providers to strike a deal with Scribd when it pivoted from being a document hosting site to a subscription platform in 2013, and the 100,000+ titles it injected into its catalog must have played a crucial role in convincing consumers that the service was worth their consideration. Scribd must have wanted it very much as well, as Coker also revealed that under the deal Scribd paid out 60% of the list price of each ebook to authors whenever a reader read over 30% of the content. Now, as Scribd said in its own statement, it has over 1 million titles in its catalog, many of which secured under too generous terms, and it needs to rationalise this in order to make the service more financially sustainable. It is also just possible that with a large quantity of romance content now in its catalog, Scribd wanted to take this as an opportunity to refine and curate what it presents from the category in a more discriminating manner.

Both Coker and Scribd's statements suggest that it's not only self-published authors who are affected. Publishers who provide romance content to Scribd may also see a dip in their title count and a concomitant loss of income, though so far none have gone on the record to say how substantially they are affected. Amazon has never disclosed under what terms it acquires content from traditional publishers for KU, so it's impossible to say whether it has also changed the way it pays them for participating in the ebook subscription platform.

Whichever way we look at this situation one thing is clear. Self-published authors benefited disproportionately from the initial rise of ebook subscription services. These services needed content, and was prepared to pay self-published authors reasonably well for the privilege of making it available to readers on an all-you-can-eat basis. They participated wholeheartedly, scenting rewards, even as publishers held back. Then, as it emerged that (in some categories at least) that the all-you-can-eat model led to binge consumption it was inevitable that indie authors should shoulder a disproportionate amount of the fallout as ebook subscription platforms sought to recalibrate their services.

What this says about the long-term sustainability of ebook subscription we will only know when the unluckiest of them runs out of venture capital money. But this situation does suggest that even though ebook subscription services would have never got out of the starting blocks without the support of indie authors, they might need to reduce their reliance on them in order to convert themselves into sustainable businesses.