Tips for the right data strategy

Data is coded employee knowledge

A data strategy also determines how companies encode knowledge about their products, services, processes and business models. This makes solutions possible that also allow automated decision support. Let’s briefly go back to our online optician: In order to sell glasses online, a lot of optician’s expertise has to be encoded so that the customer doesn’t make a serious mistake when configuring their glasses. The optimal size of the progressive lenses depends, among other things, on the prescription and the lens geometry. In order to successfully sell glasses online, this experiential knowledge of opticians must be encoded in the product data, and the various responsibilities (procurement, production, eCommerce) must maintain, connect and use this data.

A knowledge graph captures the meaning of the data and plays a special role in identifying and networking the data: Dave McComb’s three-layer knowledge graph model expands an external two-layer view of schema classes or data instances on the other hand. McComb introduces a middle level that plays a hybrid role and refers to these three levels as concepts, categories and data.

In a very practical way, Katariina Kari, Lead Ontologist at Inter Ikea Systems, and her team introduced such a knowledge graph. We are based on this example, but transfer it to the online optician example.

• The top layer contains the central concepts, for example “frame” with “properties”. The number of concepts is in the hundreds. They are closely coordinated and subject to strict central governance.
• At the middle level, the categories, “color” is defined as a property with the characteristics “tortoise” or “havana”. The number of categories is in the thousands, but the categories can be divided thematically and the individual thematic areas are defined by appropriate specialists.
• McComb calls the bottom layer data, and this layer includes what is being colored, for example the bridge of a glasses frame. The number of entities in the data layer is potentially in the millions. The data plane is broken down into areas, each of which is under the control of the domains. The principle of federalism can also be seen particularly well here.

The integration of the categories and especially the data into the entire landscape takes place via reference to the higher levels, so that networking is possible. For example, all frames can be linked to the bridge color Tortoise. For example, similar products can be suggested in the eCommerce system using similarities.

Elements of the data strategy correspond to data mesh principles

The currently much-discussed Data Mesh concept from Zhamak Dehghani, Technology Director of the IT consulting company ThoughtWorks, is nothing other than the concrete manifestation of a data strategy. This sociotechnical concept is based on the four principles of domain ownership, data as a product, self-service data platform and federated governance. We relate this concept to the four key aspects of identity, bitemporality, networking and federalism.

• Domain ownership: This principle states that responsibility for data is not borne by a central data team, but rather in the domains in which it is created. In concrete terms, this means: The team that is responsible for a subject matter end-to-end is also responsible for the data that is created in connection with this subject matter.
• Data as a product: Collecting, processing and providing data is not an end in itself, but – like every product – must create value for its user. But this also requires strategic planning, a suitable product-market fit and the marketing of the respective data product: Data products focus on the data consumer and their needs, but also balance the different wishes of different consumers. The form of a data product, for example as an API, as database access, or as a visualization, depends on the needs of the consumer, and different data products can be created from the same data for different needs.
• Self-service data platform: In order for the product teams to be able to provide their data products quickly and efficiently, they need appropriate tools, essentially a production and distribution route for data products. Ideally, these tools should be interconnected in such a way that it is also easy for consumers to network different data products. “Self service” – or perhaps better put, “following the principle of subsidiarity” – means that the data owners are independently able to offer data products. Contrary to the name “data platform”, it is equally a question of the available infrastructure and the organizational structure to structure the teams in such a way that this independence can be achieved.
  In terms of complexity, this principle represents the biggest hurdle to implementing a data mesh approach. Not because the corresponding data platforms are not available, but because the balance of competencies within the organization must be rebalanced accordingly.
• Federated governance: To create added value, the data mesh approach emphasizes locally owned data products. According to our points mentioned above, added value arises precisely in the networking of different domains, in the relationship between data producers and consumers. There are also areas that are dictated by external regulations in terms of security, data protection, etc. that cannot be regulated locally by the data owners. There must also be overarching structures and guidelines that govern how data is organized and used in larger contexts. The federal principle of subsidiarity applies: Similar to the interaction between municipalities, states and the federal government, decisions are made at these institutional levels, whose competence is just sufficient. If the individual, the smallest group, or the most cost-effective institutional level lack the skills, a correspondingly higher authority takes action.

Identity, bitemporality, networking and federalism in a data network

Depending on the business requirements and complexity of the data flows in a company, a data mesh can represent the most sensible implementation of a data strategy. All too often, emphasis is placed on the technical rather than the sociological side. But we also see that the four principles of Domain Ownership, Data as a Product, Self Service Data Platform and Federated Governance provide little concrete guidance: What does a data product contain? How does it relate to other data products? What should a self-service data platform enable?

6 reasons: Why you fail at data-driven

Here we come back to the four key aspects of a data strategy: identity, bitemporality, connectivity and federalism. These key aspects focus the data strategy on specific points and can, for example, give structure to the implementation of a data mesh:

Which identities are exposed in the data products? Which data products must reference common identities to enable networking? Do data products only have to be implemented “for the moment” or to look forward or back – keyword bitemporality.

And the question that looms over everything is: Who has the skills to identify entities? Competence means both the professional, technical and design knowledge as well as the generally recognized mandate to design the corresponding information spaces.

The data mesh approach explicitly relates the federal principle to governance, i.e. to administration including the design of administration. We go beyond this with our understanding of federalism and explicitly understand the design of the data spaces: The creation and maintenance of the concepts, categories and data in a knowledge graph is also organized as a federal structure: For the top layer, the concepts, there is one central design necessary. The category level can be broken down and implemented locally. In particular, different sub-areas of the second level can be managed by different teams. The data layer then truly arises locally in the domains and is subject to the respective owner of a data product.

Data strategy requires culture

In recognition of Peter Drucker’s “Culture eats strategy for Breakfast”, culture is also a mandatory requirement for a successful data strategy. (Corporate) culture includes the intangible foundations of an organization’s creative services.

With regard to data culture, for example, the question of the design of federal structures also arises: Does an organization emphasize central responsibility or local responsibility? Do federal levels also correspond to hierarchical levels? Are decisions also escalated via managers or are competent, i.e. decision-making, committees put together in a different way? How is the decentralized competence of the domains balanced in comparison with centrally provided platforms, which can be used with the lowest possible learning curve for the users from the domains, but which require considerable effort to do so.

Pragmatic – step by step to the North Star

Companies rethinking their data strategy should develop a north star but then proceed very pragmatically. The North Star represents the target image that is being sought: Do you want to increase efficiency, improve products or services based on findings from existing data and open up new business areas? If the goal of a data strategy and corresponding initiatives is not clear, then its implementation is doomed to failure. Only when the direction is clear can practical steps lead to success.

The organization can be changed, for example to set up federal governance structures, to realize central control of the top ontology layer and to adapt and improve it in interaction with the domains. The domains must be enabled to implement data products independently, with central definition of the policies that must apply to everyone, for example with regard to identity and access management. And here, in creating a platform – planned or emergent as a result of only loosely coordinated initiatives to reduce communication overhead – the data strategy approaches the classic IT strategy, especially in relation to cloud architectures.

Conclusion: Make well-founded decisions with a data strategy

Competitiveness through innovation requires a well-thought-out data strategy. By focusing on the key aspects of identity, bitemporality, networking and federalism, companies can unlock the potential of their data and make informed decisions.

7 data strategy trends: Is your data strategy still sustainable?

It’s not just about collecting and analyzing data, but about creating a culture of data-driven decision-making. You need the ability to strike a balance between centralization and decentralization. A core element of our society, federalism, becomes a structuring element.