Each database will differentiate itself and shine for specific workloads. A case in point is QuestDB, which deals very well with heavy ingestion (high cardinality / out-of-order data). This lends itself well to use cases around financial tick data and IoT data.
Usually they either build their own cloud offerings (and distribute them via AWS/GCP/Azure/Alibaba/etc. marketplaces) and/or cater to the old-school audience that doesn't trust managed cloud services (with enterprise licenses, support, consulting, etc.).