The value that engineers bring to the data stack is undeniable. They play a crucial role in ensuring data is primed for insights. However, the immense value and necessity of involving Domain Experts in the data supply chain, not merely as consumers but as contributors, cannot be overstated.
IT and Engineering have distinct focuses within the data supply chain. Their work aligns with the overarching goal of preparing data for insights. Yet, there's a specific persona that is interested in enhancing the intrinsic value of the data itself. Consider the analogy of a Netflix movie director. While Netflix's ability to deliver content globally, start streaming in seconds, and function across devices is impressive, the content—the movie—remains the star. It requires directors, writers, and editors to refine the content to the point where it captivates the audience. Similarly, one might ask, who is the "director" for your customer data?
Wasn't it always the plan to equip Domain Experts with tools they could use independently?
Before the advent of CluedIn, the team developed a Web Content Management System aimed at empowering content creators without IT involvement. This begs the question, what is the data stack's equivalent? Is there a reluctance from IT and Engineering to relinquish control over the data stack? Certainly, there are numerous reasons, some akin to the notion that sometimes it seems easier to do the work oneself rather than delegate.
Operational systems alone aren't the solution.
Even the most pristine CRM system, when merged with other datasets, may reveal conflicts that need resolution. Who is tasked with this? IT? Engineering? Currently, they handle it, but this approach is problematic.
What do you lose when domain experts are excluded from data management?
Excluding domain experts from master data management introduces significant risks to an organization's data integrity and operational efficiency. Without their specialized knowledge, the data managed by IT and Engineering may lack the nuanced context and real-world application that domain experts provide. This disconnect can lead to data that is technically accurate but practically unusable, resulting in poor decision-making and missed opportunities.
Moreover, without domain expertise, the data may fail to comply with industry standards or reflect the latest market developments, rendering it obsolete. The absence of domain experts in the data lifecycle also increases the likelihood of data silos, as the insights and needs of different business units are not fully integrated. Ultimately, this exclusion can compromise the strategic goals of the organization, as data fails to serve its ultimate purpose of driving informed business actions and outcomes.
What changes when Domain Experts are integrated into the data supply chain?
Initially, there will be friction due to the historical separation of IT and business within the data stack. The integration point—where business input meets IT standards—will need to be established. Furthermore, data will transition into products, complete with Product Managers, Roadmaps, and Project Planning.
They will tackle conflict resolutions, and engage with other data product teams to address issues, facilitating data integration, linking, and matching. They will adhere to Product Release Cycles.
Here are some successful strategies we’ve already observed:
- Enrich company data with every possible unique identifier to simplify linking across domains, such as Tax Numbers, DUNS, PermIDs, etc.
- Elevate values to their highest fidelity to enhance "usability," such as converting dates to full UTC and standardizing phone numbers.
- Identify and eliminate duplicates.
- Incorporate additional data to construct hierarchies and improve company data perspectives.
- Enhance data accessibility by managing glossaries that clarify the distinction between a Customer and a Prospect.
Is there an analogy that demonstrates this concept?
Indeed, source control is a prime example. It epitomizes the balance of centralized and decentralized control, embodying all the necessary elements for scalability, though it is not without complexity. This example demonstrates that the data silo challenge is solvable; the industry simply hasn't fully realized the solution yet.