Google BigQuery for analyzing large ML datasets using SQL
Updated December 20, 2019

Google BigQuery for analyzing large ML datasets using SQL

Anonymous | TrustRadius Reviewer
Score 8 out of 10
Vetted Review
Verified User

Overall Satisfaction with Google BigQuery

Google BigQuery is used for data warehouse as a ML analytics engine company wide specifically for consumer behavioral analytics with data streams coming coming out of website as well as internal data sets.

Pros

  • It is easy to create and then execute machine learning models in BigQuery using SQL queries using BigQuery ML. Everyone knows SQL.
  • Google BigQuery is fully serverless/cloud based and can be up and running in few hours without need for any specific coding or integration if your data is already is Google Storage.
  • Google BigQuery executes the SQL statements very fast and can can be used for real-time analytics especially if you use Google infrastructure ( GCP).

Cons

  • Google BigQuery is great for large data sets where you need a familiar SQL interface but it is still slower than running the same SQL query on RDBMS, assuming your data is mostly structured.
  • It is expensive if you have a lot of data that needs to be queried each time the query is run due to the license metrics used in Google BigQuery.
  • Some of the SQL operations like table join are not optimized and can be slow compared to a full database.
  • Being server-less, fully managed cloud server, Google BigQuery has a positive impact on the business in terms of amount of setup time and deployment resources needed to analyze a data set.
  • Positive impact on ROI due to reduction in CapEx and OpEx needed to provision a data warehouse upfront.
  • Positive impact on ROI due to no improvement in the speed of analyzing consumer data using Google BigQuery in real-time and proactively take action when/if needed based on the results.
Google BigQuery needs minimal setup to get it up and running while Amazon Redshift and Oracle Analytics Cloud need moderate expertise and time to load a data set and run a query. Hadoop (open source) and its commercial version Cloudera do not provide a full out of the box solution for data warehousing and need additional components and installs. Databricks is a smaller vendor and does not come into picture if you are already an Oracle or a Google shop (=using their cloud, DB, et al.)
Online documentation was readily available and it was easy to connected with the product management team for Google BigQuery during Google Cloud Next 2019 event. We didn't have to open a SR/ticket through the usual support channel to get our issues resolved!

Do you think Google BigQuery delivers good value for the price?

Yes

Are you happy with Google BigQuery's feature set?

Yes

Did Google BigQuery live up to sales and marketing promises?

Yes

Did implementation of Google BigQuery go as expected?

Yes

Would you buy Google BigQuery again?

Yes

Google BigQuery is very well suited if your data is large and already in Google Cloud/GCP where the data itself is not simple structured data. It is less suited if you have well-defined data sets that may or may not exist in Google Cloud. Google BigQuery is also less suited if you have to analyze the data on a regular basis since the cost of accessing compute and storage adds up considerably.

Google BigQuery Feature Ratings

Database-as-a-Service
9.5
Automatic software patching
10
Database scalability
10
Automated backups
10
Database security provisions
10
Monitoring and metrics
7
Automatic host deployment
10

Comments

More Reviews of Google BigQuery

  1. Home
  2. Database-as-a-Service (DBaaS)
  3. Google BigQuery Reviews
  4. User Review of Google BigQuery