To me, three biggest things that led to reduced adaption of hadoop in-house
1. Cloud computing - hadoop was (and probably is) an expensive platform to manage in-house
2. Docker and Kubernetes
3. Relational databases and SQL will take to you 95% there
