Estimated reading time: 11 minutes
Table of contents
What is a data driven enterprise?
A data-driven company maximizes data value as a product by treating its internal data as an enterprise asset and making data-driven decisions that are backed by technology, processes, and people. Data becomes the differentiating factor that propels enterprises forward acting as a competitive differentiator. In the most effective data-driven organizations, all business decisions are made using some form of data analytics. In a data driven company, senior executives use data to make strategic decisions about the future of the business. Middle managers use data to improve operational efficiency and bottom line results.
Why do we talk about data as a product?
Taking a product approach to data assets enables an organization to derive value from data in a sustainable way. It also allows for the different stakeholders involved in data decision-making to be more effective and efficient in their roles. Data Product Management (DPM) is an area of product management that specializes in the development and delivery of data products. A data product is a software application or service that makes use of data to provide a value-added feature or service. The DPM discipline encompasses the entire product lifecycle from ideation to delivery and post-launch operations. DPMs work with data science, engineering, and business teams to bring data products to market (internally or externally).
Ultimately, data products provide businesses with digital products that help them grow and compete more effectively in today’s digital economy.
How is a Data Product related to a Data Mesh?
A data mesh is a modern framework for creating data products. It allows businesses to quickly connect and process data from a variety of sources. This makes it possible to create data products that are more accurate, reliable, and up-to-date than ever before.
This also helps businesses avoid the common pitfalls associated with traditional data management approaches. For example, data silos can form when data is managed in isolated systems. This can make it difficult or even impossible to access the data when it is needed. Data mesh helps businesses avoid these silos by providing a unified platform for data management.
What are the Fundamentals of Creating a Great Data Product?
Creating a great data product requires more than just having access to data. businesses must also have the right tools and processes in place to make use of that data.
Key attributes for creating a great data product include:
- Build an Easy-to-Use Data Product
This means having a user-friendly interface that makes it easy for users to find and understand the data they need. - Build a Comprehensive Data Product
This means having a data product that has all the necessary data and analysis for the specific domain of the data product. - Build an Accurate Data Product
This requires having processes in place to ensure that data is regularly cleansed and updated. - Build a Timely Data Product
This means having a system in place that allows users to access the data they need, when they need it. - Build an Actionable Data Product
This mean data for its own sake is waste. Data needs to provide actionable intelligence. - Build a Secure Data Product
This means having security measures in place to protect data from unauthorized access or misuse. Creating a great data product requires a commitment to security at every stage of the process. - Build Useful Documentation
This also includes having clear instructions and tutorials for using the product.
By following these principles, businesses can create products that are both useful and valuable to their users. Creating a great data product requires a combination of technical expertise and business acumen. By understanding the needs of their customers and having the right tools and processes in place, businesses can create data products that drive real value. When done correctly, data products can help businesses make better decisions, improve customer service, and increase sales. Creating a great data product starts with understanding the fundamentals.
Build an Easy-to-Use Data Product
The benefits of easy to use data products are numerous. First and foremost, they make it easy for customers to find the data they need. They are also comprehensive and accurate, which means that customers can rely on them for accurate information. Additionally, they are timely and actionable, which means that customers can take immediate action on the data they receive.
Build a Comprehensive Data Product
When it comes to data products, comprehensiveness is key. A comprehensive data product allows businesses to create valuable comprehensive insights. To build a comprehensive data product, businesses need to connect and process data from a variety of sources. This includes internal data sources, such as databases and CRMs, as well as external data sources, such as social media and web analytics.
Build an Accurate Data Product
Accuracy is critical when it comes to data products. After all, if the data is inaccurate, it won’t be of any use to businesses or their customers. To build an accurate data product, businesses need to ensure that they are collecting high-quality data. They also need to ensure that this data is processed correctly and stored in a reliable data warehouse.
Build a Timely Data Product
A data product is only as good as the timeliness of the data it contains. To build a timely data product, businesses need to ensure that they are collecting real-time data. They also need to ensure that this data is processed quickly and made available to customers in a timely manner.
Build an Actionable Data Product
An actionable data product is one that provides customers with the data they need to take action. To build an actionable data product, businesses need to ensure that they are collecting data that is relevant to their customers’ needs. They also need to ensure that this data is processed in a way that makes it easy for customers to find and use.
Build Useful Documentation
One of the best ways to create valuable documentation for a data product is to use the Documentation Quadrants as a pattern. The Documentation Quadrants can be used to help businesses document all aspects of their data product, from the basics, to the nitty gritty details. The four quadrants are:
The Tutorial Quadrant, which is Learning-Oriented and includes Quickstart Guides and Recommended Learnings like Reading Lists and Training Lists.
The How-To Guides Quadrant, which is Problem-Oriented and includes Troubleshooting Guides and IT Operational RunBooks.
The Discussion Quadrant, which is Understanding-Oriented and includes Chat, QnA Bots, FAQs, and Stack Overflow.
The Reference Quadrant, which is Information-Oriented and includes API Docs, Data Catalogs, Service Catalogs, and Data Lineage Reports.
Enterprise Architecture
Business Architecture
Data Product Value Propositions
During the next phase, the software product is designed for the needs identified above. Data products deliver value propositions. Sometimes it is necessary for data products to support a specific problem. Data products that feature a value proposition can nudge people to provide information, whereas the analytics of information can lead to data products that help decision-making. In some circumstances, a used car platform could include a data product called “Predicted Used Car Resale Value” which would help in determining selling costs.
Three types of data products
There is a three-type data-product in my opinion: data as a service, data-enhanced products, and data as insight.
Setting up Your Data Product Team
Data products are sometimes part of larger service platforms with numerous data product managers, data scientists, data analysts, data engineers, and data consumers that handle the different aspects of developing and operating the data services. These data teams need to excel in cross functional collaboration and have a product mindset.
Information Systems Architecture
Architecture for Data Products
The architecture for the enterprise data products should take in the following considerations:
Data quality: It is important to ensure that the data being used for the data product is high quality, accurate, and up-to-date.
Data processing: It is important to ensure that the data is processed quickly and correctly so that it can be made available to customers in a timely manner.
Data storage: It is important to ensure that the data is stored in a reliable data warehouse so that it can be accessed by customers when needed.
Data security: It is important to ensure that the data is secure and confidential so that only authorized users can access it.
Data sharing: It is important to ensure that the data is shared with customers in a way that is easy for them to understand and use.
The technical assets the enterprise data products should have include the following:
Code: Any software that includes data pipelines that is used to utilise, transform and serve upstream downstream data and code that receives upstream data. API code provides access to information and semantic schemas along with code that enacts the security, compliance and policy responsibilities of the user
Data and Metadata: Analyses historical data related to a domain. Data could be in any format like structured, unstructured or semi-structured
Models: Logical, conceptual and mathematical models that explain how data is related. Models are used to make predictions.
Training datasets: A set of data used to train machine learning models
Data Mesh, Data Products, and Data Science
In reality, Data Products belong to the larger picture: the Data Mesh. Data mesh isn’t merely a technology, its platform is the solution for organisations to untapped its data capabilities in order to create better AI/ML knowledge sharing applications. Please take the time to read Zhamaks most influential book Data Mesh – Delivering Data-Driven Value at Scale. This will not be a quick summary but we want to reaffirm that data is an integral component of data mesh architecture.
Conclusion
Data products are an integral part of the mesh, and play a important role in allowing companies to utilise their data for better performance. Data products help organisations to make predictions and better decisions by providing high quality data that is processed quickly and accurately. In order to set up a data product team, it is important to have a product mindset and excel in cross-functional collaboration. The technical assets that the enterprise data products should have include software, code, data, models, training datasets, and metadata.
FAQ
Data products are tools, applications, and datasets that use data to improve decision making processes.
Common data products include Salesforce’s Einstein algorithm, which specializes in customer predictive statistics, and finance terminals such as Bloomberg Terminal.
To make a data product, you need to develop an understanding of the business problem that you’re trying to solve, and take a product management approach to solve that problem. After that, you’ll need to choose the right technology platform and tools to build and deliver the data product.
A data mesh is an architecture that uses many small, distributed data services to give users easy access to the data they need. Data meshes are designed to be scalable, fault-tolerant, and easy to use.
Data products are designed to be easy to use and accessible to as many people as possible. By making data products readily available, organisations can increase the number of people who are able to use data to improve decision making processes. This means reusing expensive to develop business models, business logic, deep learning findings, and artificial intelligence systems to provide better insights.
A “Credit Score” for a Financial Services company, a “Service Maintenance Predictor” for a Manufacturing company, a “Churn Model” for a Telco company, or a “Product Recommendation Engine” for an E-commerce company are all examples of data products.
RegulatedCloud offers consulting services to help you get started on turning data (raw data or dirty data) into the right data.
Born and raised in Houston, Texas; Jacob brings 20 years of technology experience, with previous roles as a chief information officer, principal architect, cloud engineer, and full-stack software developer; while producing technology that manages $1 trillion worth of financial transactions. He currently has over 40 IT certifications.