Creating trust in data starts with asking a range of questions. Where is the data coming from? Is it a trusted source? Does it provide the details we expect and need? Is it the right quality? Is there a clear process and trusted source to enrich the data? Satisfying these questions gives us the confidence to deliver data for use.
These questions were relatively easy to answer in a pre-cloud and SaaS world of structured data. However, today’s universe of devices, external platforms, data volumes, and complexity makes data governance and trust a high and expensive bar to clear. Even as we’ve moved from the Hadoop era to cloud platforms like Azure, Snowflake and Databricks, most data isn’t easily accessible, fully understood, or governed. This means most data isn’t valuable to the business, mainly because people don’t trust it.
The growth in data transformation is aimed at breaking this cycle. While some of those projects deliver focused use case ROI, the promised land of organizational-wide change still alludes most. CDOs and CIOs have investment fatigue, focusing mainly on maintaining forward momentum and ensuring governance.
Into this complicated universe comes GenAI, with a top-down mandate to immediately apply it to everything in sight.
Creating the Trust Needed for GenAI Success
The age-old data trust and quality issues take on a new dimension with GenAI. Initial experimentations clearly show that GenAI’s full potential comes when blending structured and unstructured data. While GenAI-powered insights can positively impact specific roles, value chains and user journeys, they lead to new questions. For example, how do we know each source is trustworthy as we blend data? Which data type provides the most valuable information for a specific question?
Users need to trust the insights they receive. With all the market noise around hallucinations, open vs. closed models, and potential job automation, employees are fully aware of the risks of using bad data. The workforce’s well-founded reticence could stop enterprise GenAI usage in its tracks.
This is why trusted data foundations are more important than ever. Building one starts with embracing a simple reality: data means different things to different people. This naturally leads to thinking of data as more than a generalized resource to solve a point-in-time question. Successful organizations will adopt a Data as a Product approach that considers each user’s unique needs as part of their data foundation work.
Tailored data products, with their inherent understanding of how data will be used, make it easier to apply data quality, privacy, security, and governance for each specific experience. Also, instead of a big-bang approach, it enables the incremental build-out of the data foundation or the extension of current data foundation principles and policies to new data sources. This operating model creates alignment so information is accessible, available, trustworthy, and cost-effective. The result is higher user confidence and increased adoption, creating a virtuous value cycle based on powerful experiences with trusted data.
GenAI has the potential to enhance every product, experience, and process. But without a trusted data foundation, adoption will stagnate and GenAI’s promise will go unrealized. With trusted data foundations enabling a Data as a Product approach, organizations can set the conditions for GenAI to enhance everything across the business.
