For many years, Teradata was the undisputed leader in large-scale data warehousing. Banks, insurers, and telcos built their most critical systems on it. Today, the market is very different. Cloud-native databases such as Snowflake, BigQuery, and Databricks have set new standards in elasticity and simplicity.
The question is: what role will Teradata play in this new world?
What Teradata Offers with Hybrid
With VantageCloud Lake, Teradata presents a hybrid architecture. The idea is simple:
- Shared-Nothing MPP (classic Teradata): predictable performance for workloads designed around primary indexes.
- Shared-Data MPP (cloud object store): elastic compute nodes reading and writing object storage.
The optimizer decides which approach to use. In practice, this means customers can continue to run legacy workloads without rewriting everything, while gradually expanding into the cloud.
For large organizations, this is attractive. It reduces migration risk and protects investments made over decades. It also allows cost optimization by keeping hot data in Teradata storage and moving colder workloads into cheap storage.
The Strengths of the Hybrid Model
- Continuity – Existing queries, reports, and ETL pipelines keep running.
- Lower risk – No “big bang” migration to a new platform.
- Flexible economics – Expensive storage only where needed, cheap storage elsewhere.
- Skills – Teradata experts remain useful, new teams can work on the cloud side.
It is, in many ways, a bridge to the cloud.
Where Hybrid Falls Short
However, hybrid is not cloud-native. It inherits much of Teradata’s complexity:
- Two execution models instead of one.
- Data movement between storage layers.
- Ongoing need to manage primary indexes, skew, and statistics.
- Slower innovation pace compared to pure cloud players.
- Talent attraction challenges, as most engineers now learn dbt, Airflow, Snowflake, and Databricks.
Hybrid buys time. It does not remove the long-term pressure to modernize.
Market Signals
Teradata’s revenue trend has been under pressure in recent years, even as cloud clients increase. Cloud-native competitors are expanding much faster. This makes it unclear how Teradata will maintain or expand its overall market share.
A Possible Future: Teradata as the Next Mainframe
Here is one possible scenario I see:
- Teradata remains important in industries like banking and insurance.
- Critical, compliance-heavy workloads stay on Teradata.
- New projects and analytics increasingly move to Snowflake, Databricks, or BigQuery.
In this world, Teradata plays a role similar to the mainframe today: still used in a large percentage of global banks, absolutely critical, but ultimately a niche.
Conclusion
I like both Teradata and cloud-native databases. The hybrid model is a sensible intermediate step for large enterprises that cannot risk a full migration today.
But in the long run, I am not sure what Teradata’s strategy will be to hold onto significant market share. The most likely outcome may be that Teradata becomes a specialized system, still vital for certain workloads, but no longer at the center of the data ecosystem.