Building products and applications in a world dominated by cloud providers (AWS, Azure, GCP) is hard. For Startup Founders in a space that they dominate, this should be what keeps you up at night. The risk for the clouds to build a "me too" product and brand you as a "middleman" happens all too often.
Why should a customer, who uses one of the cloud stacks, buy your product when the cloud provider offers something similar that's "good enough?" Granted, your product might be better but it's considered both a capital and operational expense. Why take on that risk when the cloud provider gives you a competing product and lets you build what you want as purely an operating expense?
This is the siren song of the cloud providers, consume what you need and turn it off when you want. Why buy something else when the cloud provider has a plethora of tools in an ever-expanding toolbox?
Take, for example, AutoML, once the darling of the machine-learning world. The first entrants in this space did relatively well until the cloud providers acquired or built their own "me too" products. Customers could simplify and replace any subscriptions or licenses from a vendor (you) for what their cloud provider offers.
The immediate clue to the cloud providers taking your market share is churned customers and flagging revenues. By the time you realize that the cloud providers are slaughtering you, your marketplace will be awash with the blood of your other competitors. This is what professors W. Chan Kim and Renée Mauborgne coin a Red Ocean in their book "Blue Ocean Strategy." The large market participants (aka cloud providers) eat up all the smaller competitors, consolidating the market space.
Here's a great primer on Red vs Blue Ocean markets:
The answer to surviving in a Red Ocean is to find a Blue Ocean, one where the competition isn't and can be made irrelevant. It's creating a market where none exists. So how do you do that, exactly? Do you need to pivot? The answer is to start at the beginning.