There are two discomforting truths within digital transformation across our key industries; energy, utilities, and manufacturing.
- Digitalization PoCs are commonplace. Real ROI isn’t.
- Billions are invested in cloud data warehouses and data lakes. Most data ends there, unused by anyone for anything.
At the heart of this data-driven value dilemma lies a confluence of challenges, ranging from the technical (How can we best organize our diverse and fluid data universe?) to the operational (How can we create new information products and services?), to the financial (How can we treat data as an asset?), to the human (How can we improve data literacy and ensure digital solution adoption in the field?).
Read also: DataOps: A transformative new approach to data ROI
To avoid boiling the ocean, we will focus on what is perhaps the most fundamental question all fellow Chief Data Officers and other digitalisation executives need to consider as their Northstar — and in doing so, we will find ourselves on the right path to debunking both discomforting truths above. Which sets of my IT/OT/ET data universe are creating measurable business value right now? (Beyond their source business system of origin of course.)
To get to data value, focus on data governance, not data storage
Let's face it, over the past decade, data governance hasn’t enjoyed any of the hype and executive attention its sibling cloud data storage has. Quite the contrary, data governance continues to draw an association with data management hygiene such as access management and security on one hand; and please-don’t-bore-us-with-cryptic-IT-back-office-jargon such as metadata and lineage on the other.
Learn more: The data liberation paradox: drowning in data, starving for context
The tide is however turning fast, catalyzed by the meteoric rise to fame of data quality — or lack thereof — as the promoted root cause for why despite billions invested in cloud data warehouses and data lakes, most data ends there, unused by anyone for anything. An inescapably alarming fact is that “data has no value unless the business trusts it and uses it,” as summed up by Forrester.
Solving the data quality challenge is however not as straightforward as filling cloud data warehouses and data lakes have been. Investing millions into another doomed MDM project — only this time for the cloud — is equally erroneous. Instead, adopting a data product-centric mindset, along with a DataOps practice to create and manage data products, is needed.
Use data products to deliver data quality (and real value) to business
Data products are not business applications or solutions using data that you, alongside your SI partners, deliver to your business. For the sake of clarity, let's refer to these simply as data-driven business applications (mobile apps, low code apps, dashboards, etc.) What we mean by data products is your data itself being “packaged” into a true self-service product experience for the data products’ customers.
A useful analogy is to think about providing businesses with SaaS instead of code snippets in Git. Similarly, data products are MUCH more valuable than their raw data inputs because they come with intuitive user guidance (descriptive metadata and context) and guaranteed quality in the form of an SLA (incl. data product owner to contact for support). Most importantly, data products are built — similar to SaaS — with actual data product customer needs in mind. They are not raw data sets post-QA.
Data product usage shows where true data value lies
Similar to that in SaaS, a sustained increase in data product usage is an unquestionable metric of customer value.
Moreover, as you are measuring data product use and not raw table queries, you obtain valuable insights into how your data landscape needs to be refined further, to be even more purposeful and valuable for business. You equally obtain a common language to use with businesses to articulate their data product needs. For example, having access to all pressure and temperature changes along a specific pipeline as opposed to a dump of raw time series ingested from a particular local historian.
Becoming digitally native is by no means a short or straightforward journey for any industrial enterprise. For me as Chief Data Officer, being able to answer which sets of my IT/OT/ET data universe are creating business value — and where and how — need not be a headache; it can be an accelerator.
Learn more: Drain your data swamps with data-products-as-a-service model
Implementing Industrial DataOps to collaboratively develop, manage, and operationalise data products is how you get there. Once there, progressing to direct data monetization using data products is suddenly at your fingertips...
Did this article cause some reflections? We’d love to hear from you and your experiences.
Author: Petteri Vainikka. Published originally at: https://www.cognite.com/blog/data-products-and-data-value-measurement