14 ms·
Amazon's Kindle Direct Publishing is a dystopian nightmare
Imagine this scenario: You get pulled over by a robot cop. The robot immediately terminates your driver's license forever for driving an 18-wheeler without proper certification. You point to the Mini Cooper you were driving when the robot pulled you over. The robot ignores you.
Eventually, a different robot shows up, listens to your defense, sees the Mini Cooper, and watches you start the car with your own keys. There is no 18-wheeler in sight. The new robot says it will get back to you in 5 business days.
Two weeks go by, and a third different robot tells you your appeal has been denied for "driving in such a way that creates a negative experience for other drivers". All the stuff about 18-wheeler certification is never mentioned or acknowledged again.
You appeal again and again. All your appeals are heard by yet more robots, who always uphold the original robot's decision, regurgitating the same bland phrase about creating a negative driving experience. THE END
This is basically what has happened to me with Amazon's KDP. I published my paperback through Ingram Spark and my Kindle eBook through KDP. My eBook passed review and was for sale for a few days. I claimed the paperback on my author central page, where it appeared side by side with the Kindle version. I was told to wait a week and the two versions should link up automatically (so they appear as the same book). If not, I should send an email to Amazon support.
Then the eBook was blocked, and a day later I got this in an email from KDP's Content Review Team:
> During our review, we found that the following book(s) creates a misleading customer experience by impairing customers' ability to make good buying decisions.
> Land Without a Continent: A Road Trip through Mexico and Central America
> Items that can cause a misleading customer experience include:
- Similarity of the contributor name to another author
- Similarity of the title to a previously published book
- Similarity of the cover to a previously published book
- Cover text or images that do not accurately represent the contents of the book
- Title or subtitle that do not accurately represent the contents of the book
- Similarity of the description to a previously published known work
I replied to their email pointing out that there is no book with a remotely similar name, description and author name as my book. I even did a reverse image search on my cover, and the only hit was my book on Amazon. It seems obvious to me that their AI-powered fraud detection system hallucinated that I plagiarized my own paperback.
The next day, they terminated my account. I appealed. I got an automated reply saying to give them 5 business days to review my case. That stretched into two weeks. Finally, I got an email stating my appeal was denied with only this message:
> We found that you have published titles with misleading metadata (including cover), which creates a negative customer experience.
It seems like they picked the only one of the six bullet points that couldn't be easily disproven, and ran with it. I continued to email and managed to get a few more informal appeals, all denied with the exact same vague nonsense message. I even emailed jeff@amazon.com.
I had finally given up, when I got an email from Robert from their Executive Customer Relations Team. Hooray a human being (maybe)! A day later, Robert upheld the termination with the exact same vague reason above and some semi-belligerent language about how this was my last appeal.
Here's plenty more reading:
https://writersweekly.com/angela-desk/and-even-more-complaints-about-amazon-kdp-kindle-direct-publishing
https://www.trustpilot.com/review/kdp.amazon.com
https://medium.com/@peteylao/my-cover-story-aka-how-i-managed-to-get-my-kdp-account-suspended-terminated-and-finally-fd3631e84aba
https://judylmohr.com/2024/02/02/my-amazon-nightmares-2024/
- beej71 2y agoI sell through KDP and this is a good heads up. Sorry it happened to you; when it eventually happens to me I'll bail to Lulu and never look back.
- shafyy 2y agoDoes anybody have experience with using a service similar to Lulu in Germany/EU?
- RCitronsBroker 2y agohttps://indie-autoren-buecher.de/selfpublishing-blog/buch-veroeffentlichen-als-selfpublisher-anbietervergleich https://indie-autoren-buecher.de/selfpublishing-blog/buch-ve... here’s a German article discussing KDP alternative, maybe that’s of help to you.
- shafyy 2y agoThanks!
- jasfi 2y agoWhen companies can't tell you what they believe you did wrong, and point to a vague list of things it could have been, they fail.
- deely3 2y agoSo, when Google and Amazon will fail?
- packetlost 2y agoGoogle will fall very slowly as it gradually degrades in public image and engineering quality. It weathers the oncoming storm better than most others simply because it did not invest heavily in cloud as a product and source of revenue. Instead it's mostly a way to make marginal dollars to offset some of the upfront and operating costs of infrastructure it already had and needed to continue to build. Amazon will probably go much quicker as they primarily make their money from providing cloud services, subsidizing their delivery network and storefront with money made from the high margin cloud products. Microsoft is similarly betting very hard on cloud and AI, moving its bread and butter business services (ActiveDirectory, Office Suite, Windows) to subscription model SaaS products while spending billions on AI infrastructure and integrations. When the AI bubble pops, it takes most of the Fortune 500 (primarily in the Fortune 10) with it. VC investment dollars for fast growth startups dry up as portfolios prefer more conservative, stable growth businesses. Consumer spending nosedives, primarily effecting Amazon as the razor thin margins of its delivery network start getting eaten into by less volume without reduction in op-ex. It increases margin on cloud products to compensate. Nvidia retains a sizable chunk of its AI driven valuation. Microsoft, who bet hard on cloud and AI hurts, badly. They dramatically raise prices, starting with their government cloud customers, but public cloud quickly follows suit. Businesses look to alternatives, shifting towards owning their own "private clouds". Oxide computing is eating good. Oracle, who has been building data centers for AI workloads, suddenly has excess of capacity freed up by the drop in investment for AI. They either invest in their own low-margin, primarily self managed cloud infrastructure or go all in on being the American Hetzner, selling U-s and racks populated with (by this point) last generation hardware that businesses are moving towards as their IT op-ex budgets get slashed and are forced to cut costs by decreasing infrastructure spend at the cost of reduced quality and shouldering the risk of failed hardware and poor security posture. SysAdmin and DevOps roles start focusing more heavily on fundamentals instead only particular IaaS platforms and APIs, similar for dev roles. Meanwhile, Microsoft is on the brink of collapse. It sells off its Office suite to Intuit, Xbox turns into a certification platform for PC-prefabs and gets sold to Tencent. Windows continues to exist, but struggles as the new generation entering the workforce primarily uses browsers and chromebooks and businesses shift harder towards Linux and ChromeOS as the workstation OSes of choice because of costs and ease of management using free, self-managed platforms or small up and coming competitors to ActiveDirectory. Amazon and Google struggle but survive. Microsoft gets broken up and follows in IBMs footsteps and fades out of relevance.