Better data conversations start here.

Or “Why Data Landscape Maps help organisations create a shared understanding of information that everyone can use, not those of us passionate about data”

One of the biggest challenges in data governance and management is creating something that everyone can use. Too often, data artefacts are designed for specialists, understood by specialists, and ultimately ignored by everyone else. That's where Data Landscape Maps (DLMs) are different.

They create a trusted visual artefact that becomes the basis for conversations across an organisation, helping business experts, data professionals and leaders develop a shared understanding of the information they rely on every day. Rather than disappearing into technical documentation, the DLM becomes a reference point that people can trust, share, refine and use to create better data outcomes.

As we discussed in the Connected podcast (or you can watch us in person on Connected You Tube Channel,) that's exactly why we’re increasingly creating DLMs as starting points for all our work. They are simple enough to be understood by anyone, yet powerful enough to support governance, architecture, data quality and change initiatives. While many data tools focus on documenting information, a DLM focuses on creating a simple shared understanding of your data world.

When everyone is looking at the same picture, conversations become clearer, assumptions get challenged, and organisations are far better equipped to make informed decisions about their data.

If you've listened to the latest Connected Data Podcast episode, you'll have heard Rav and I talking about Data Landscape Maps (DLMs) and why they've become such an important part of our approach to data governance and management. We spend a lot of time describing what they look like and how they work, but the challenge with a primarily audio format is that DLMs are fundamentally visual tools.

That's why we've included the driving test DLM example alongside this blog. While we talked through it in the podcast, this is your opportunity to actually see what we mean when we talk about topics, relationships and data domains, and understand how a simple visual can support much bigger conversations about data.

So, What Is a DLM?

At its simplest, a DLM is a visual representation of a data domain. Before we explain what it is, it’s important to understand what it isn’t!

We're not creating detailed database designs. We're not documenting every field. We're not focused on systems or technical implementation. Instead, we're creating a shared view of the information that exists within a particular area of an organisation.

For example, in higher education we might create a landscape for the student domain. In manufacturing it might be inventory or production. In finance it could be budgets and transactions.

The goal is simple: create something that both technical and non-technical people can understand, discuss and contribute to because they can see their “data world” as an intuitive non-technical diagram.

What makes them work?

One of the biggest challenges in any organisation is that people naturally understand their own data really well, but often have limited visibility of what sits outside their immediate area of responsibility.

When teams come together to build a Data Landscape Map, something interesting happens. People begin to see connections they didn't know existed.

Comments like these are incredibly common:

• "I didn't know we collected that data."

• "We have something similar in our team."

• "I didn't realise you were maintaining that information."

• "We've been solving the same problem twice."

The map creates a common language around data without requiring people to learn technical modelling techniques. That's often when the shared understanding and real value starts to emerge.

A simple example.

The driving test example we discussed in the podcast is a great illustration of this.

A first version of the landscape – as we’ve shown above - might contain topics such as:

·       Person

·       Instructor

·       Examiner

·       Test

·       Assessment

·       Vehicle

·       Test Centre

·       Result

Within those topics sit smaller groups of information such as contact details, licence information, assessment criteria, dates, locations and outcomes.

What's interesting is not whether the model is "right". In fact, during the podcast we challenged our own assumptions and evolved parts of the example as we talked through it, and that’s exactly the point.

Data Landscape Maps are collaborative by design because they encourage discussion and land in an area of shared agreement. On the way, they help people uncover assumptions, duplication and pockets of uncertainty. In some cases, the conversation generated by the map is more valuable than the finished diagram itself.

DLMs and DG, a perfect fit?

Although DLMs are useful for lots of things, one of the reasons we use them so frequently is that they provide a natural entry point into data governance.

Traditional governance conversations can sometimes feel daunting. Telling someone they've been assigned as a data steward and need to follow a governance framework is rarely the most engaging introduction. Especially if done by email!

However just asking people to explain what data they hold, use and transform, is a far gentler way to scope out a data domain and the responsibilities within it.

Once the map exists, it becomes far simpler to identify:

·       Data owners

·       Data stewards

·       Areas of responsibility

·       Known issues

·       Critical data topics

·       Gaps and overlaps across teams

Meaning prospective stewards can see the context of their role rather than simply being assigned one. How to onboard a steward is a whole different topic we will get to soon!

And then what?

One of the biggest misconceptions is a DLM is only useful for governance activities. Our experience shows this is just a jumping off point for many other uses. Here are just a few.

Data Architecture (especially for new projects)

Enterprise and data architects often use landscapes to assess the impact of major business or technology changes. Whether an organisation is replacing an ERP platform, introducing a new application or redesigning a business process, the map helps identify which data areas may be affected, and set expectations about the quality of that data.

Data Discovery

Trying to find data in a large organisation can be surprisingly difficult. A landscape provides a starting point by showing what information exists, who is responsible for it and where people should begin their search. It may also highlight which is the master system for data, what reports are available and – crucially – who to talk to when help is needed.

Data Quality

One of our favourite use cases is visual data quality management.

Topics can be highlighted to indicate whether information is considered trusted, requires caution or needs improvement.

This creates a much more accessible way of communicating data quality than lengthy reports or spreadsheets. It can also be trended over time. So the DLM can track change and – hopefully – progress. The same applies for data risk, information compliance, etc.

Prioritisation

When multiple data challenges compete for time and investment, a landscape helps everyone see where improvement efforts are likely to have the greatest impact. It becomes much easier to agree priorities when everyone is working off a trusted artefact.

So we’re done then? Not quite.

One thing we stress is that a Data Landscape Map is never truly complete. As organisations change, systems evolve and people learn more about their data, the landscape evolves too.

They are at a level of abstraction that simple (even system) changes can be easily dealt with. But the fidelity and context improves with each interaction.

The more people use the map, the more valuable it becomes. Additional metadata, governance information, business definitions and quality assessments can all be layered on over time, transforming a simple visual into a powerful organisational asset.

Anything else?

We're particularly excited about the next step in the evolution of Data Landscape Maps.

Many of the DLMs we create today are developed and managed in Lucidchart, but we've been working on a new visualisation and data portal platform that allows those maps to be imported and enriched with additional information.

The result is a far more interactive and manageable experience, making it easier to connect data governance, metadata and organisational knowledge directly to the landscape itself.

We'll be sharing much more about the platform in the coming months and will be running a webinar to demonstrate what's possible.

In the meantime, we're looking for organisations that would be interested in joining our testing group and helping shape the future of the platform. If that sounds like something you'd like to be involved with, we'd love to hear from you.

What are you waiting for?

If you're curious about Data Landscape Maps, start with the driving test example attached. Don't worry about creating the perfect model. The objective isn’t to make it “right”, it is to make it an agreed starting point for whatever you want to do next.

Get a few people in a room, pick a familiar process, and start drawing lines and boxes.

You might be surprised by how much you learn about your organisation's data in the process.

And if our experience is anything to go by, you'll discover that the conversations around the map are every bit as valuable as the map itself.