- 1. Align data points with the company’s business strategy
- 2. Document how and what type of data is being captured
- 3. Define the process for data lifecycle and storage
- 4. Create a process for data governance
- 5. Communicate business semantics throughout the organization
Hearing complaints from teammates in your organization about forecasting issues? Is your idea of a sales opportunity different than your co-worker’s? Are big decisions made on “gut feel”? Chances are your organization needs a data strategy. A data strategy is a plan for how a company will collect, store, manage, and use its data to achieve its business goals. A well-defined data strategy can help codify processes, prevent communication errors, and improve efficiency, profitability, and customer experience.
A data strategy is a plan for how a company will collect, store, manage, and use its data to achieve its business goals. A well-defined data strategy can help companies improve their efficiency, profitability, and customer experience.
Here are 5 actionable steps an organization can take to define their data strategy:
1. Align data points with the company’s business strategy
The first step is to understand your company’s business strategy and key performance indicators (KPIs). Evaluate your company’s current data estate - identify where data comes into your organization and how it naturally moves through as business processes are conducted. Pinpoint what data points are most valuable in relation to KPIs.
Data Point Alignment Examples
- Construction: A construction company might focus on data points that are related to project timelines, costs, and materials. This data can be used to improve project efficiency and profitability.
- Health sciences: A health sciences organization might focus on data points that are related to patient demographics, medical history, and treatment outcomes. This data can be used to improve patient care, develop new drugs and treatments, and conduct research.
2. Document how and what type of data is being captured
Once you have identified the key data points that you need to collect, document how it will be captured and stored. This includes identifying the different sources of data, the methods of data capture, and the data storage systems that will be used. The data will be both structured and unstructured. Structured data contains information like alphanumeric text, financial numbers in a database record or text and numbers in a spreadsheet that are saved in an organized structure. This type of data requires less storage compared to unstructured data. Some examples of unstructured data are images, video, audio, flat files (PDF, Excel, Word), social media posts, chat logs that have varying attributes and are saved without any organization.
To understand how to store pertinent data, establish data provenance and single source of truth lineage. For instance, if it’s data for a construction company, data may originate in project management, accounting, or CRM software that’s duplicated from one software to the next. Which software database will be considered the source of truth for a data point? Another example from the medical billing industry may involve how patient data will be securely captured where it will be stored to comply with HIPAA - is a private on-premises server required or is there a HIPPA-compliant cloud storage solution available? These are examples of questions to seek answers to when documenting data capture processes and storage.
3. Define the process for data lifecycle and storage
Your data storage process should define how data moves from one system to another once it’s captured. For example, you might have a process for transferring data from your CRM system to your data warehouse on a daily basis. AWS, Microsoft Azure, Google Cloud, and Oracle Cloud Infrastructure all have a multitude of object storage and database options. Open-source databases like MySQL and PostgreSQL are ideal for structured data storage. MongoDB, an open-source NoSQL database, is best-suited for storing unstructured data. These open-source databases can be hosted on any of the aforementioned cloud providers.
The data lifecycle should describe how long different types of data are stored and how they are disposed of when they are no longer needed. For example, you might keep customer data for seven years after the customer last interacted with your company. Often data security and lifetime protocol are established by compliance regulations like NIST or PCI. Verify if your organization is required to comply with these types of regulations.
4. Create a process for data governance
Data governance is critical for maintaining data integrity over time. To start designing a data governance process, establish annual, semi annual or quarterly check points. Often the time frame of these check points can be based on standard business cycles in your industry. For instance, a seasonal industry like construction in the northern U.S. may slow down during the winter, reducing the number of data transactions. The checklist below covers items to address during data governance check points:
- Review data access permissions and security policies and decide if the right people have the right access to the right data at the right time.
- Inspect data quality to determine if data values, formats, and accuracy have changed significantly from one check point to the next, looking for data that deviates from a known pattern.
- Are data points that are being captured and processed still relevant and aligned with business KPIs? As a company’s industry and business environment evolve, new data points may emerge that need to be included in the strategy. Conversely, some data points may not be relevant.
5. Communicate business semantics throughout the organization
Finally, to create momentum around the newly defined data strategy, make an effort to socialize the important data points, or business semantics, throughout your company. Business semantics are the definitions of each of the critical data points that make up your data strategy. Communicating these data points and their related definitions ensures the entire organization is using the same language when talking about data. This has a ripple effect of creating business efficiency and reducing mistakes.
One way to communicate business semantics is to create a data dictionary. A data dictionary is a document that defines the data points that you are collecting, their definitions, and their formats.
Following these 5 steps for creating a data strategy and implementing the strategy will position your organization to unlock value and efficiency with your data. Still unclear on where to start? ⇨ Contact us to understand how to being creating a solid data strategy