This expression is used so often, but do we take its meaning for granted? Let’s go back to basics and explain the importance of being data-driven.
Sometimes, it’s hard to see the picture when you’re inside the frame.
This simple truth applies to many things — perhaps none more than the domain jargon that we‘ve become so intimate with that it becomes borderline impossible to assume an outside-of-the-frame perspective to its meaning. In a recent conversation with a CTO, he directly asked the simple question:
“What exactly do you mean by data-driven EA?”
This to-the-point question demanded reflection from our team. Have we somewhere down the line just ’skipped over’ clarifying the critical distinction in the context of EA? Has the workflow and architecture of being data-driven vs. drawn up manually — or being based on fixed frameworks resulting in pre-defined visualizations — been overseen by us? Living up to our mantra that seeing is believing, we wanted to share the conversation here, and the whiteboard drawings that followed.
Data-driven EA can best be understood as a commonly seen tech stack as illustrated below. Next, let’s discuss each layer in the stack, and then return to how this approach to EA is fundamentally different from traditional approaches.
It’s All About the Data
At the risk of sounding sentimental, this warrants a short trip down memory lane to Ardoq’s founding idea — resulting from the frustration of a senior enterprise architect consultant working on the migration of a banking payment system for a large Nordic bank that had gone through several acquisitions.
“I was working in MS Visio, drawing up illustration after illustration of the payment system landscape, integrations, and relationships to business processes for the client, and every time I presented the latest drawings for discussions — in particular to new stakeholders — someone pointed out yet another new system, integration, or process that was missing in the current drawing.
Each time, I found myself back at the drawing table, tediously redrawing diagrams. This led me to document each component, relationship — and their respective attributes — in Excel first and then move from there to Visio periodically.
However, it soon became clear that Excel was not the right tool for the task. Nor was it understandable for anyone else, and I couldn’t find what I was most critically looking for — gaps in the data.
The challenge was therefore a data analytics one. A challenge that would be unlocked by being able to analyze complex structured data at its most raw level.”
- Magnulf Pilskog, Co-founder of Ardoq
Ardoq’s disruptive approach to addressing enterprise architecture as a data analytics challenge is rapidly embraced by progressively thinking customers globally.
With enterprise IT in the cloud, embracing digitalization and data-driven decisions has never been easier. The data that EAs and CIOs care about can now easily be accessed via APIs.
We‘ve also built smart surveys to crowdsource complementary data from the entire organization for no-UI, no-new-solution-training data collection at an unprecedented scale.
Data that underpins everything within New Enterprise Architecture at Ardoq. It is also an incremental and living asset that follows the enterprise as it transforms.
Data Empowers Automatic Visualizations
Humans respond to visual stimuli better than anything else. This is why television quickly became a larger medium than radio, and why social media has seen a seismic shift from text to pictures and video.
When embracing a data-driven approach to EA, we can quickly, easily, and literally in seconds create many different visual representations of the same structured data. We also don’t need to redraw anything when the data changes, but changes in data — ‘at the source of truth’ as we like to call it — automatically trigger updates in all visualizations everywhere. This not only saves thousands of hours of time, but also removes the possibility of human error in the updating of visualizations.
For data visualizations, Ardoq comes with out-of-the-box, solutions including process flows, heat maps, spider charts, sequence diagrams, swimlanes, relationships, integrations, dependency matrices, bubble charts, tree maps, and strategic roadmaps. Not only that exhaustive list, but the tool allows anyone to create their own custom visualizations.
Filter and View From Different Perspectives
Last but not least, we arrive at the top of the data-driven EA stack hierarchy with filtering and dynamic perspectives. Because every visualization is based on data, any attribute pertaining to any graph component or graph relationship can be used to filter the automatically produced visualization.
By leveraging Ardoq’s graph data architecture, you can enable visual analysis to be provided on the same visualization from different points of interest or perspectives. For example, with a click of a mouse visual analysis can be changed from answering roadmap to answering risk.
Backed By Real Data
We live in a data economy. Data powers most of today’s most successful digital enterprises, and is poised to disrupt many traditional industries that have not yet realized the transformative power of data within their business.
Our research shows that the single most important thing to management is meaningful insights, backed by real data, delivered in a way that is easy to understand, and helps businesses make better decisions. Data-driven enterprise architecture can answer complex questions with real data, at the speed of digital, ideally suited to deliver on these desires.
Why not dive deeper to learn more about why New Enterprise Architecture is about real-time augmented intelligence, not architecture.
We want to engage and connect stakeholders across the organization. We listen to customer pain points, deliver solutions, and write about them.