Utilizing the Medallion Architecture to Optimize Your Domo Credit Usage

In today’s data-driven world, organizations are inundated with mountains of data that need to be processed to derive actionable insights.

More and more data storage, processing and visualization tools are available almost everyday. Which tools your company uses drives what an efficient data architecture is for your business.

Domo is a full back-to-front BI solution, able to store, process, and visualize data all in one platform.

When utilizing Domo and the consumption model pricing, the question then arises of where do I process my data? Do we bring all the raw data into Domo to have as much granularity as possible but use more Domo credits? Or do we pre-process the data, bringing in just clean, report ready data to reduce credits?

Processing data before loading it into Domo reduces credit consumption and in turn, saves costs. However, are there other costs you’re not considering with that approach? Are you losing value from your data that could cost your company a lot more than the extra credits would have?

With Domo, the best data architecture for your company finds the balance between not losing any value from your data, and not creating a complex, unmanageable workflow. A helpful way to think about this is with the medallion architecture, a data design pattern Databricks coined.

In this blog, we’ll talk through the medallion architecture approach and how it can guide your data architecture in Domo.

What is a Medallion Architecture?

In a medallion architecture pattern, you look at three layers of processing data, with the goal of incrementally improving the quality of data as it flows through each layer.

The bronze layer is the raw data integration; data is stored without processing or transformation.

The silver layer is filtered, cleaned, and augmented data; tables are transformed into a more usable format. Note that silver layer data should not be aggregated, it should output clean, augmented, granular data.

The gold layer data is refined to meet business and analytics requirements. The main step to gold layer data is often aggregation, but it can include additional transformation such as specific filtering and or calculated fields.

A medallion architecture is a simple data model that guides how and where data should be transformed. It outlines a logical progression of data cleanliness and report preparation. For your business, you can think through what data sources you have and what tools you use, such as Domo, Databricks, Power BI, etc. and use the medallion architecture as a framework to build your optimal data workflow.

Optimizing Domo Credits and the Medallion Architecture

For legacy Domo contracts, you typically had a number of connectors, rows of data storage, and a number of user license. Within those limits, you could connect and build anything. In the medallion architecture framework, most businesses brought in bronze layer data into Domo. Starting with the raw data, they did all of their clean up and transformation in Domo. This method enabled data savvy teams to access their data and reports all in one system. It also can lead to more ETL maintenance and potentially redundant dataflows. Companies often didn’t have a clear data plan when they starting building ETLs, sometimes leading to the Domo data warehouse ballooning without tangible reports on the front end.

In Domo’s consumption model pricing, you’ll want to be deliberate about your data architecture. With the medallion architecture in mind, think about what layer of data it makes sense to bring into Domo: bronze, silver, or gold.

Need help optimizing your domo consumption usage? Connect with a Domo consultant today.

When finding the balance between Domo credit consumption and not losing value from your data, often the answer will be to bring in silver layer data. Connecting data that is already clean with demographic data joined in will reduce dataflows while not losing value from your data.

Silver layer data allows companies the option of having teams be self service with their analytics, or having an analytics team built reports for each department. Many report requests can be built right off of silver layer data, handling any aggregation and additional filtering in Analyzer and with beast modes. Building cards off silver layer data allows users to drill into the details they need, and reduces the need for an additional dataflow.

There are, of course, many use cases when it is still valuable to bring in bronze layer data to Domo. For most marking use cases, unless the workflow is already setup, it does not make sense to pipe the data from the various marketing sources into another system first, to then bring it into Domo. You risk losing too much value from your data and not having full control to filter and drill into the data end users need.

Bringing the bronze layer marketing data into Domo enables your teams to get the most from the data. With both sales and marketing teams often utilizing the marketing data, the bronze layer data allows each department to transform and build reports with the data as needed.

Alternatively, some businesses have data warehouse or existing workflows that could suggest connecting gold layer data to Domo. For example, if your sales team utilizes Salesforce, you could choose to bring in prebuilt reports from Salesforce instead of raw objects and fields. With this architecture, cards would be built directly off the report datasets in Domo, and should validate to Salesforce.

Conclusion

In the consumption credit structure, your data architecture should be deliberate and governance should be at the forefront of your Domo use. However, the credit structure should not dissuade you from bringing your valuable data into Domo. With the medallion architecture in mind, you can be deliberate and thoughtful about the best data workflow for your business, getting the most value from both your data and your investment in Domo.

If you’re not sure where to start, here is a few ways to optimize your Domo consumption model.

As premium Domo partner, we are strategically positioned to assess your unique use case and help ensure the optimal data workflow. Contact our team today to discuss your Domo project and how to best structure your data!