Building A Scalable Data Warehouse With Multivariate

Setting up a data infrastructure has 3-4 major components - which are as follows:

• Ingestion from Sources via ETL Pipelines (Fivetran, Stich, Panoply)
• DBT (Optional - Hevo)
• Database (Bigquery, Redshift, Snowflake)
• Visualisation Tool (Tableau, Mode)

Setting up a data warehouse - the old way

Hiring data engineers, working with custom solutions is very difficult and expensive. Moreover, there are various caps that you have to consider while setting it up - for example - some ETL companies keep pricing based on connectors - in addition to the number of rows processed. It all works out to be very expensive setup.

Each segment has the below limitations:

• Number of rows processed
• Number of source connectors (Analytics, Google Ads, Facebook, Hubspot etc).
• Data stored.
• Data edited/updated.
• Number of users - in case of visualisation tools.

We felt all this cost was too much of an overhead - especially for small businesses that are just looking to get things up and running. Most of the businesses need well over 2M rows/day. And the cost of all this - comes in thousands of dollars.

Therefore, we decided to add all components together and keep pricing fixed only on the number of rows processed - with a base cost to run essential services.

Here's how our pricing looks like:

• Pay on the number of rows that are processed per month.
• Managed DB - pricing based on the GB of data stored - No limit on queries.
• Visualisation tool - 5 users free, then $10/month/user.
• 100+ Integrations - no additional pricing on the number of integrations that you use.

Pricing Comparison:

• Over 50% lower than going with individual options.
• Marginal cost of processing and storing data is 70% lower than other solutions.
• No limits on querying your data.

Other advantages

• Delegate the data handling to one company - no need to reach out to 4 providers for individual solutions.
• Need new connectors? Tell us, and we'd build custom connectors.