This Snowflake is Unique, Fast and High Performing
May 02, 2021

This Snowflake is Unique, Fast and High Performing

Anonymous | TrustRadius Reviewer
Score 9 out of 10
Vetted Review
Verified User

Overall Satisfaction with Snowflake

Snowflake was our data warehousing solution of choice, which we migrated from Treasure Data. It is used across the whole organization. It is used as a source of data for the entire organization, powering our dashboards and machine learning solutions.
  • Computation is handled under the hood, freeing resources that would be used for maintainence of clusters
  • Handling large data scale and ingestion
  • Ability to query large volumes of data with speed due to their unique architecture
  • Snowflake UI can be clunky and breaks sometimes, which can be annoying
  • Snowflake has to be paired with the Data Build Tool (DBT) to allow for templating and macro usage. No inbuilt solution.
  • Snowflake Python connector development doesn't necessarily track popular packages such as Pandas as quickly as Pandas releases
  • Could do with better machine learning capability over warehouse tables, but I assume this is coming soon
  • Query speed as analysts need to look over large volumes of data
  • Handling of computational scaling under the hood, unlike with older solutions such as Hadoop and Spark, so data warehouse team is smaller
  • Advanced functionality and pairing with data build tool (DBT) allows for data quality tests
  • Positive impact: we use Snowflake to track our subscription and payment charges, which we use for internal and investor reporting
  • Positive impact: 3 times faster query speed compared to Treasure Data means that answers to stakeholders can be delivered quicker by analysts
  • Positive impact: recommender systems now source their data from Snowflake rather than Spark clusters, improving development speed, and no longer require maintainence of Spark clusters.
Our initial data warehousing solution was Treasure Data. We had issues with the costly pricing model, which would be exhorbitant if we want to hold our data in memory and query using Presto. As a result, some heavy lifting was done in Hive (managed by Treasure Data); unfortunately, this is also very slow.
As a result, dashboards are updated slowly, and analysts found it hard to deliver insights to stakeholders in a timely manner.
Snowflake was a blaze compared to Treasure Data. It solves two problems: computation can be easily scaled with respect to the data volume, and this is handled by Snowflake, and query speeds are much faster than Treasure Data.

Do you think Snowflake delivers good value for the price?


Are you happy with Snowflake's feature set?


Did Snowflake live up to sales and marketing promises?


Did implementation of Snowflake go as expected?


Would you buy Snowflake again?


Snowflake is a powerful warehousing solution and suits companies with large scale of data. It helps with fast querying of data, and there is little need to manage computation, since it is managed for you.

However, it does require a dedicated team and an upfront cost in setting up an structuring the warehouse. Some solutions such as AWS's Redshift or GCP's Bigquery could be a better alternatives if is already within the AWS or GCP ecosystem. Bigquery in particular has a low upfront cost and a better pricing model.