Data Product Management has been a regular topic of discussion among Data Science, Engineering and Product Management teams over the last few years, particularly when Data Science products and Machine Learning are involved.
The role of the Data Product Manager has many similarities to that of a Software Product Manager in that a keen understanding of customer business requirements is crucial. There are however some key differences to their respective responsibilities and the skill sets needed.
In what business environment does a Data Product Manager usually navigate in?
It is fair to say that Machine Learning dependent products impact our daily lives. Social media platforms (Linkedin, Facebook, Twitter), Google, Uber and Airbnb have all developed highly sophisticated ML algorithms to improve the quality of their product.
Today, Data Science products are by no means the chasse gardée of the top tech companies however. They have also become a common feature in a variety of enterprise related domains such as predictive analytics, supply chain management, crime detection, fraud, high staff turnover prevention to mention but a few.
Data Product Managers are often called for when Data Science Products are involved, in other words when the core business value in the spotlight depends on Machine Learning and Artificial Intelligence.
So what does a Data Product Manager do?
Again, the role of the Data Product Manager is analogous to most Product Management roles in that it is geared towards developing the best possible product for the customers/users. That remains the key focus for the Data Product Manager.
There are, however, some subtle differences when it comes to the remit of the Data Product Manager…
The population range that the Data Product Manager caters for is often wide and can include Data Scientists, Data Engineers, Data Analysts, Data Architects and even developers and testers. Such a diverse pool of expectations requires a solid understanding of each of these fields in order for the Data Product Manager to understand the use case for each stakeholder, not to mention strong people skills to navigate through these different universes unscathed.
To demonstrate the diverse range of skills involved in this new role, the ideal Data Product Manager will have a broad understanding of Machine Learning algorithms, Artificial intelligence and statistics. He will have some coding experience (enough to dip his toes in if needed), be good at math, understand the Big Data technologies… and have second to none communication skills.
The Data Product Manager can even be assigned the responsibility of centralizing access to Data at the enterprise level*.
Here, he might be asked to come up with new ways to manage, collect and exploit data in order to improve the usability and quality of the information. This part of the job may involve choosing a suitable Data Management software to centralize and democratize access to the data sets for all parties mentioned above, breaking down silos between teams and facilitating data access for all.
They may then choose a Data Catalog platform with a powerful knowledge graph and a simple search engine…such platforms do exist ;).
*How, in this instance, does the role of the Data Product Manager differ from that of the Data Steward you may ask. After all, isn’t it up to the Data Steward to curate, manage, handle permissions and make the data available to the data consumers? One way to consider the distinctions between the two roles could be to see the Data Steward as the data custodian of the data of the present and the Data Product Manager as the custodian and innovator of the data of the future.