How to Identify Your First Data Products

As your business grows, so does the demand for data. This almost always leads to more data models, tables, and dashboards. This can make it nearly impossible to understand the health of key data assets from the lens that matters to end-users. For example, you may wake up to a dozen test errors but have no visibility into whether key business processes are impacted or who should be notified.

This is why you need Data Products. Think of Data Products as grouping your data assets into constellations that make sense for the business. Everything else then follows—from defining ownership to testing and monitoring and measuring SLA and uptime.

What is a Data Product?

At its core, a data product is a set of data assets that serve a specific need. Unlike a data pipeline that simply moves data from point A to B or a table that’s a piece of a bigger puzzle, a data product is designed to be reusable, and valuable to end-users—just like any other product in a marketplace.

Data products can take different forms, including:

• Datasets (e.g., clean, curated customer 360 data models)
• Metrics & KPIs (e.g., a key revenue metric)
• APIs & services (e.g., a real-time fraud detection API)
• Dashboards & reports (e.g., a Looker report for sales performance)

We recommend you think about Data Products as either producer or consumer products.

Producer Data Products are read from operational systems (e.g., API or SalesForce data) and owned by data or platform engineers. Encapsulating them in data products provides an easy-to-understand getaway for downstream teams to escalate issues upstream without grasping the full complexity of the internals of the data products.

Consumer Data Products directly expose data to consumers and read from other data products or assets (e.g., CLTV or Marketing Attribution Model), often owned by data analysts or scientists. Encapsulating these in data products gives everyone a good understanding of issues’ direct impact and which stakeholders should be notified.

How to identify Data Products in your business

Identifying your Data Products is not about labeling everything but focusing on what drives value. If you can identify the most critical business processes your data team supports, those are most often the data products you should identify.

If you’re unsure where to start, following these two steps is a good starting point

1. Start with the business needs

Before thinking about schemas, tables, or models, ask

• What business processes that rely on data are most critical?
• What decisions or processes depend on data?
• Who are the end-users, and what do they need?

A data product should solve a problem or enable a workflow—not just exist for the sake of it.

For example: If marketing frequently asks, “What’s the lifetime value of a customer?” or sends lifetime value data to an ad provider through an API call, and you build a curated dataset to answer that reliably, you likely have a Data Product.

Data Products identified this way will often be consumer Data Products and owned by data practitioners with domain expertise.

2. Look for upstream bottlenecks

Data products can be a great way to make core data sets easier to understand and use by downstream data teams. Look for data where

• Data assets that have many downstream dependencies
• Various teams spend a lot of time cleaning and preparing the data
• Data that depend on data engineers for manual extracts

These will often be your producer Data Products.

The number of Data Products is highly dependent on your company’s size and complexity. Some companies have dozens or hundreds of data products while others have just a handful. Start small, by identifying a handful of your most important data products, and then build out from there.

You should also consider the granularity at which you define your Data Products. The data team at Aiven started with high-level products such as Sales and Marketing but realized they needed to go a step deeper to have the most impact.

“If the Marketing data product has an issue, that may be fine. However, if the Attribution data product within Marketing has an issue, we must immediately jump on it. This is the level of detail our data products need to be able to capture.” - Stijn, Data Engineering Manager

Making Data Products work

Identifying data products is just the beginning—you also need a framework to manage them so that they’re discoverable and reliable.

This includes:

• Ownership & severity – assigning who’s responsible and their relative importance
• Discoverability – making it easy for teams to find and use the data
• Monitoring & testing – ensuring reliability over time

For the full blueprint for how to build reliable Data Products from start to finish, check out our guide