Analytics Strategy & Data Culture
What makes a good data strategy?
Data Culture
Over the past months and years, I have discussed with various customers how the desire for more BI self-service can be implemented in everyday life. CubeServ has suggested establishing the Business Analytics Platform as a prerequisite. This makes the data usable across all technologies.
Can our users work productively with the data if we simply grant them access? Probably not. It’s not enough for our goal of making good decisions based on facts.
What makes a good data strategy? There are five dimensions in particular that need to be considered holistically in the strategy:
- Data: Data is relevant, high quality, and the various inventories are integrated.
- Enterprise: How is analytics organized to enable widespread adoption?
- Leadership/Inspiration: Does the data strategy fit with the lived company culture? Are facts and clean data preparation used and demanded to make decisions in the company? As Gary Loveman, COO Harrah’s, often puts it, “Do we believe it’s true? Or do we know?”
- the (measurable) goals (for analytics)
- as well as support for controllers, data scientists, and (casual) users (i.e., all analysts).
Zhamak Dehghani’s Data Mesh concept is a revolutionary (in my eyes) approach to decentralized data architecture. For Data Mesh – the construction of a data network, Ms. Dehghani establishes four principles:
- Domain Ownership
- Data as a Product
- Self-Service Data Platform
- Federated Governance.
The concept was first introduced in 2019 and has since spread very quickly. For example, at Zalando, which relies on the Data Warehouse Cloud with the Space concept, among others, to make Data Mesh a reality.
The simple premise of data meshing is that business units should be able to define, access and control their own data products.
The idea is that stakeholders in a particular area understand their data needs better than anyone else. When business people are forced to work with data engineers or data scientists outside their domain, providing the right data to the right data consumers at the right time is time-consuming, often error-prone, and ultimately ineffective.
If the expression of data as data products is done in the respective domain (business unit), is the central analytics team redundant? Probably not! The framework to make such data networking effective and efficient is a real challenge. The data basis can only be provided in an automated and standardized way. Permissions that ensure the protection of personal data as well as company secrets, a meaningful data model, a data catalog as well as a data marketplace remain major challenges. Even if the responsibility for the data products lies in the respective domains, the analytics team should provide coaching on analytical methods and the various standard analytics applications.
The shaping of an appropriate data strategy can only be done individually. We would be happy to support you in finding the right one for your company and planning its implementation.
You have the choice:
Take advantage of our expert knowledge on analytics strategy
