SQL databases are reasserting dominance by integrating flexible data types like JSONB. PostgreSQL achieved 55.6% developer adoption in the 2025 Stack Overflow Developer Survey, jumping 7 percentage points from 48.7% in 2024 — the largest single-year expansion in database history. This evolution allows developers to combine the rigorous, transactional integrity of relational databases with schema-less flexibility of document stores.
The JSONB Advantage
Relational databases, particularly PostgreSQL, have enhanced their feature sets with native JSONB (Binary JSON) support. This allows a single database row to contain both strictly defined columns for critical data and flexible JSON objects for less structured information like application logs or user configurations.
The hybrid approach directly counters the polyglot persistence architecture popular around , which advocated using different databases for different jobs — PostgreSQL for transactions, MongoDB for document storage. JSONB enables developers to maintain strict ACID compliance for essential data while accommodating unstructured data within the same system.
Concrete Adoption Numbers
PostgreSQL’s 7-percentage-point surge in 2025 represents ecosystem consolidation, not gradual growth. Professional developer adoption reached 58.2%, creating an 18.6-point gap over MySQL’s 39.6%. Meanwhile, competitors stagnated: MongoDB declined 0.7%, Oracle grew 0.1%, and MySQL increased just 0.2%.
The trend extends beyond surveys. Supabase, a PostgreSQL-based platform, grew from 1 million developers in 2024 to 4 million in 2025, achieving a $5 billion valuation in October 2025. The company now creates 2,500 new databases daily, demonstrating PostgreSQL’s evolution from database to platform.
Why Vector Search Accelerated Adoption
The pgvector extension became a catalyst. Released in May 2023, it enabled in-database vector search for AI applications—embeddings, RAG systems, semantic search—eliminating the need for separate vector databases. By late 2023, every RAG tutorial and AI startup demo defaulted to Postgres + pgvector instead of dedicated vector databases like Pinecone or Weaviate.
Real-world impact was immediate. Multiple companies publicly ditched Elasticsearch for ParadeDB (pg_search + pgvector) in 2024-2025, running BM25 + vector hybrid search in one query without separate clusters or data synchronization.
The Tradeoffs Remain
While architecturally powerful, SQL hasn’t eliminated its inherent complexities. Developers must still manage connection pools, write migration scripts, and navigate scaling challenges that can be more complex than scaling document stores. The perceived “friction” of SQL—defining schemas and relationships upfront—is increasingly viewed as a feature encouraging disciplined system design, but it remains a real operational consideration.
Market Consolidation, Not Replacement
The industry is moving toward balanced middle ground. SQL databases are not replacing NoSQL wholesale but absorbing their most valuable features. This suggests database consolidation where developers rely on a single, powerful system for wider task variety, simplifying architecture and reducing operational overhead.
This aligns with the Lindy Effect—the longer a technology survives, the longer it’s likely to persist. Having endured decades through multiple technological eras, SQL’s longevity demonstrates foundational strength, adapting by absorbing features like JSONB rather than being replaced.
What Remains Unclear
Specific industry-wide adoption rates for hybrid SQL-JSONB models beyond developer surveys are not yet widely published. Comprehensive performance benchmarks comparing native JSONB operations at scale against specialized document databases for various workloads remain limited. Which specific industry sectors are leading this consolidation also remains unclear beyond general fintech and SaaS adoption.
Follow Hashlytics on Bluesky, LinkedIn , Telegram and X to Get Instant Updates
