Great Start but Not at Scale
Updated October 14, 2019

Great Start but Not at Scale

Anonymous | TrustRadius Reviewer
Score 4 out of 10
Vetted Review
Verified User

Overall Satisfaction with Treasure Data

Treasure Data is used to ingest analytics from our frontend and backend services, as well as to import production tables. It serves as our Data Warehouse. This allows reporting to stakeholders to be done.
  • Managed service: don't need to scale up the infrastructure by ourselves.
  • Simple to use UI.
  • Good support from Treasure Data when we have problems.
  • Presto split hours limitation is a problem for us—we've essentially scaled quite quickly so the current plan is becoming expensive for us.
  • It's quite slow, but that's due to lack of computational resource.
  • When there are a large number of tables to list, the UI becomes slow and unresponsive.
  • Treasure Data was quite useful in the initial stages of the data infrastructure here at Canva and has helped in answering important business questions.
  • We're not getting too much value these days, since the next tier plan isn't economical for us.
  • We've integrated Embulk with TreasureData to get daily snapshots of some production tables.
  • TreasureData has been used to run queries to validate foundational tables.
In terms of query speed and performance, Google BigQuery and Snowflake offer better performance at a lower cost. BigQuery's pricing on just the data scanned rather than cost of computation is far more attractive than Treasure Data's current model. We've selected Treasure Data initially as these services were still in their infancy. Also, we didn't have a mature data engineering team at that point in time.
It's quite well suited when you're a smaller company, not so when you have heaps of events coming in like Canva. It's also useful when there is a lack of resources to manage your own infrastructure. The easy to use UI, in general, makes it easy for analysts, and sometimes, stakeholders to write SQL queries and reports.