Diary of a Data Product Manager: A case for data strategy, product definition and stewardship.

I want to share some lessons learned as a data product manager who has managed data products for the past few years. This blog will be the first part of a blog series documenting my reflections, challenges encountered, and recommendations on resolving them to deliver business values.

Data without processing and enriching it with subject matter experts' knowledge has no business value. The need to process and analyse the vast data created by business applications and processes has created opportunities and challenges for data and analytics specialists.

The first gap I have seen across projects is the need for a well-defined data strategy. The starting point for a data product manager is to work with the key stakeholders to create the data strategy for the business. A good data strategy will align with the business strategy, vision and objectives with a roadmap to deliver it.

The team is only set up for success with a data strategy and vision. Key business objectives and results (OKRs) should inform a good data strategy. In successful data projects, a good data strategy involves creating OKRs for each use case, which is then put in as a roadmap item.

The data products will build upon the data strategy. A data product differs from a data asset, and understanding their distinction is very important. Within the data team, a data product manager must identify the key capabilities of data products. This exercise is best done with the engineering team to ensure that the team understands what needs to be delivered.

For instance, a data product should have governance, data quality, data lineage, monitoring and alerting and usage metrics. All the features should deliver business value in line with the data strategy. Generally, a table in the data lake, metadata information, and visualisation dashboards are classified as data assets.

If the above two points are well defined, the data team will have a good foundation to build upon and will likely deliver value to the business successfully. Data products will differ for every business depending on the data generated by the application, end users, operating systems and processes. However, the principle remains the same. Having clear capabilities and definitions is important.

The last point I will highlight in this blog post, arguably one of the most common gaps I have encountered across projects, is the need for clear ownership. With GDPR and similar regulations, data governance has become more critical; however, the practical application still needs to be improved.

Most companies have a DPO helping to ensure they remain compliant with the regulations, but data ownership through assigning stewards is often overlooked as an integral part of this. To be successful, each data product must have governance with clear business owners.

Data product stewards should be SMEs with good knowledge of the business process that led to the generation of the underlying data used to build the data product. By involving the right owners, data products can be built and developed sustainably to deliver value to the business.

Such SMEs will be key contributors to the development of business glossary and definitions of the attributes in the data product, which makes the product more useful and accelerates the development of the data catalogue, which helps with data democratisation. More on this later.

In summary, I have highlighted the need for a well-defined data strategy, a good understanding of what makes a data product and data ownership as crucial elements for data product development and lifecycle management.

In the next blog in the series, I will discuss the role of technology in data product development and management.

?

?

?