Why Data and Analytics isn’t Really About Technology
More than ever before, today’s business leaders understand the value of leveraging data to solve challenges and make more effective decisions. Gathering, organizing, and analyzing data enables teams to enhance their products, manage their supply chain, increase customer retention, improve operational efficiency, and gain an overall competitive advantage. To address these requirements, organizations prioritize the development of data-driven strategies to drive action through real insights.
But what is a data-driven strategy? Let’s start with what it isn’t.
- It isn’t just about having a data warehouse or lakehouse, or access to the technologies collectively described as Artificial Intelligence (AI). Having the technology doesn’t necessarily mean it will be used well.
- It isn’t marketing-speak about being insights-driven — radical decisions are few and far between, and to pretend otherwise would be disingenuous. In a company of 100,000 people, decision-makers will not be driven by new insights every day. So, a data-driven culture is not about data-driven insights.
A data-driven strategy is about prioritizing the context and richness of all your data and using technology to curate your data for improved decision-making.
Those of us who have been in the data and analytics space for a while know that we’ve been discussing many of the same strategies for decades. The pendulum continues to swing back and forth between centralized and decentralized approaches. While many new terms and technologies have emerged with the rise of the Cloud, we’ve been exploring similar principles, just with different names.
“Centralized,” IT-driven approaches strive to move all data into one location to create the elusive, “single version of the truth” such as the classic data warehouse. On the other hand, organizations pursuing “decentralized” architectures focus on the need for data to be consumed directly by business users. Data mesh, data fabric, and self-service fall into this camp.
But with all these new technologies and approaches there’s been a fundamental error— the focus has been on the technology rather than on the data where the value resides. Organizations have been extracting and moving data to serve new technology instead of bringing technology to their data. In doing so they’re removing the valuable business context that exists and spending copious amounts of time trying to re-create it.
With a data-first approach you refocus on the data and stop the pendulum from swinging. Rather than starting with the question of “how do I move my data into this new technology?” you start with, “what questions do we need to answer and what decisions do we need to make?” You consider what you’ll lose by removing the data from its context. The difference is subtle, but the implications are significant.
There’s a better way to maximize the value of your data to make the most impactful decisions. With SAP data and analytics solutions, you maintain the complete business context of your SAP system data with the ability to enrich your SAP landscape with any other data you need. It’s time to put business back in business data and deliver data that’s rich with context and meaning.
Interested in learning more? Check out the eBook, Back to Business: It’s time to reset the clock on data analytics, to learn the five essential capabilities of a modern data and analytics stack. Listen to the LinkedIn Live discussion, Why Data and Analytics Innovation Isn’t About Technology to hear data and analytics experts discuss how organizations can start taking a data-first approach.
Nice try ...
but there is no context IN the data.
Even in source-systems there are a lot of layers of connecting, interpreting and transforming the data to finally give it some "context".
So bringing technology to the data does not mean to have quick access to all the context.
You cannot access THE "sales order" or THE "material master data", it is always/everywhere a combination of naked tables/columns/rows/streams and it is different in every company.
Besides that, most of the time there is a huge amount of data, that cannot be accessed/processed in the source-system, with the required (analytics/business-)logic.
I agree with you when encouraging the users to think about virtual access instead of ETL and especially asking the right questions before rashly extracting data.
In terms of "context" there is no system or technology that is doing the magic for you.
You always need the users+skill(s) to prepare the data and users+skill(s) to give it a useful meaning.