The Data Fitness Regimen: Keeping Your Data Strategy in Peak Condition
An organization operates much like the human body with numerous interconnected systems that must function harmoniously. Even seemingly simple tasks, such as bingeing streamed content, involve the coordination of all 11 organ systems. A robust data system shares this complexity. Even straightforward projects can involve many contributing sources and types of data. In this article, we'll explore the steps required to construct a resilient "data body" that can deliver results on par with professional athletes.
Define your data vision.
To embark on any project, it's essential to establish a clear vision. Avoid vague objectives like "centralizing data" and instead specify the use cases your data system will support. Consider aspects such as insights, targeting, measurement, and pricing. To ensure alignment and enthusiasm across your organization, gather input from various teams early on. Remember, your cross-functional colleagues will both consume and contribute to the data.
Identify essential data types, and which are essentially "the appendix."
Vestigial parts have little or no purpose in the modern-day human body, the appendix is the most well-known example. They served a purpose at one time but didn't disappear from our anatomy (even though their purpose did).
Just as the human body has vestigial organs with no discernible function, your data strategy may contain elements that serve no purpose. For example, data that was selected or acquired by a former team member. Purging our new "data body" of vestigial parts begins by listing out the types of data we do need. Begin by cataloging the required data types based on your envisioned use cases by category.
- Zero-party data: This type centers on capturing what your users want from you. While its primary focus is on customer or user preferences, certain fields within it may prove invaluable for a broader data strategy.
- First-party data: Building goodwill internally is crucial here. While collecting first-party data may seem straightforward, the variety of sources and formats can pose significant challenges. This includes everything from website activity to purchase behavior to marketing touchpoints. Be mindful of audience fragmentation, as this is an issue that warrants attention. Ensure that every channel and device currently part of your marketing or customer engagement strategy is accounted for.
- Second-party data: Unless you have a large strategic relationship already established, second-party data engagements will likely be low priority when establishing your data system. Most such relationships involve the use of clean rooms, making it unlikely for second-party data to become a permanent fixture in your "data body."
- Third-party data: Instead, concentrate on the types of data required to fulfill your use cases that cannot be sourced internally. For instance, consider data needs like "prospects within my market area" or "demographic insights about my current customers."
Compile your data shopping list.
Like any fitness improvement plan, you may need to do some shopping at the outset. While new trainers might not be on your list, third-party data could be. When evaluating potential data contributors, pay close attention to critical metrics such as data quality, scale, and accuracy.
This is also an opportunity to streamline your vendor relationships. You might discover that you're engaging with three separate vendors when a single one can meet all your data requirements. Additionally, you may uncover contributors of "Vestigial data" that no longer serve a purpose and should be removed from your plans. This not only reduces vendor overload but can also lead to cost efficiencies.
Choose your data spine.
The data spine serves as the central linking dataset that connects all the data within your "data body," making it actionable according to your vision. While it might be tempting to designate your first-party data as the spine, it's important to recognize that the spine should interconnect all data, including first, second, and third-party data.
A well-constructed data spine should meet the following criteria:
- It should feature a persistent identifier, which can unify various disparate identifiers. While email is a commonly considered persistent identifier, it's worth noting that people often abandon email addresses over time. Consider alternatives like names with physical addresses for a more long-term foundation.
- It should have the ability to connect to a wide array of identifiers. Your list of contributing data sources may utilize various identifiers, ranging from digital identifiers like cookies or MAIDs to communication identifiers like phone numbers or email addresses. Ensure that your spine accommodates not only current identifiers but also those that may emerge in the coming years.
- It should possess the flexibility to connect individuals to households. This flexibility is crucial not only for organizations focused on household-level purchases but also for marketing channels such as direct mail and television, which operate at a household level. A spine focused on individuals will make marketing attribution a challenge.
Don't skip annual data checkups.
Once your data spine has been selected and data is flowing from both internal and external systems, your data vision will begin to materialize. At this point you may be tempted to consider the initiative complete, though organizational structures evolve, goals change, and consumer behavior continues to shift.
New demands and challenges will emerge, and individual teams may address these changes in isolation (e.g., adding a new lead source). Just as physical exams take a comprehensive view to detect medical issues requiring attention, it's essential to conduct an annual review of your entire data strategy.
Ensure that your original data vision continues to align with the organization's needs. Assess whether any changes necessitate new data requirements or the removal of obsolete data.
This is where clear documentation makes all the difference. A simple document that outlines the data vision, use cases supported, data required, and central linking spine essentially serves as your data strategy's medical records.
A well-constructed and maintained body can do incredible things. Just as a healthy body can perform incredible feats, a robust and well-curated data strategy can empower your organization to reach its maximum potential.
The views and opinions expressed are solely those of the contributor and do not necessarily reflect the official position of the ANA or imply endorsement from the ANA.
Scarlett Shipp is the CEO of AnalyticsIQ, and the driving force behind the company's innovative data and identity solutions. She brings incredible expertise from her over 30 years of experience in the big data, analytics, and computer software industries. Scarlett previously served as President and COO, where she oversaw product development and compliance, ensuring AnalyticsIQ stayed ahead of an ever-evolving regulatory landscape. Scarlett loves traveling the world, enjoying a great glass of champagne with her husband and friends, staying up to date on the latest fashion trends, and spending time with her dog, Elvis.