How to protect the integrity of your data

What is data integrity?

There are two sides to data integrity — physical and logical.

Physical data integrity refers to the accuracy and completeness of data in storage and access. Things that can compromise that are anything from power cuts, to hackers, to human error, or storage erosion. That last term is when data becomes unreadable, perhaps because it’s stored on a now-unreadable format — an extreme example would be a floppy disk.

Logical integrity is about keeping data intact and consistent however and wherever it’s used, so there’s no duplication, modification, or corruption of the information.

Maintaining your data integrity

Like data integrity itself, its maintenance has two sides. One is culture, and the other is technology, and they support and enable each other.

If you have poor data integrity, then you have a poor data culture by definition. In a good data culture, things wouldn’t have reached that stage. It’s a simple fix, though not necessarily an easy or quick one. Everyone in the firm will need to adopt new behaviours and shed old ones — identify the habits people have or don’t have that have led to poor data integrity, and encourage them to start or stop doing what they need to.

You know your firm best, so the specific approach will be up to you, but generally speaking, the project will be more successful and sustainable if you inspire genuine buy-in. Those who understand and believe exactly how data integrity will help them in their role will be more likely to put in the effort to build and maintain it.

Technology bolsters your data culture in two ways. In a small but valuable way, it helps achieve the buy-in just mentioned — by investing in software to assist the new data practices, you demonstrate your support for your teams as they make the change. In a larger practical way, the right software makes following best practice easier than not following best practice, which is the secret to making even data-apathetic teams get into good habits.

The goal

There’s no finish line — for one thing, no firm will ever have perfect data integrity. What’s more, as a project, data is more like a garden than, say, a sculpture. You don’t finish then leave a completed project, never to work on it again — you attend to it and maintain it indefinitely.

While there’s no sense of the project having been ‘completed’, there is a standard that will tell you if you’ve done a good job on landscaping and planting the proverbial garden. That standard is a Single Version of the Truth (SVT), which we discuss in greater detail here. In SVT, every data point exists consistently wherever it’s found — there are no different versions, saved by different people in different places. Whenever someone accesses firm, client, or case information, they get the definitive authoritative version, and there’s no risk of using outdated, inaccurate, or corrupted data.

If you have a Single Version of the Truth, you have data integrity. Now you have to maintain it.