One Big Bad Assumption

In business, many people have to go in with some assumptions on how things are going to work. The one that affects data the most that is really bad is this:

We have data, so it must be good.

Wrong, in most cases its poor quality. Just because data is produced as the by-product of doing business doesn't mean it is good.

You definitely don't get there by accident.

Just like when you cook you take raw ingredients and make something with it. No one eats flower straight from the can, they use it to make a cake and eat that. Same concept for data. Raw data is collected, processed and delivered in a useable form and still needs to meet certain quality standards to actually be quality and useful.

What Is Quality Data

As I've said before, all businesses generate data through the course of conducting business. However there is a different between any data and quality data that is useful for identifying opportunities, streamline operations, understanding customers and so forth.

Quality Data has 9 dimensions:

• Complete. The data is complete without missing values.
• Consistent. The data is consistent and doesn't say the same thing 15 different ways.
• Accurate. The data is accurate, true and reflects reality.
• Timely. The data is up to date and available at the appropriate time. For example, you don't need to wish a customer happy birthday, 6 months late.
• Valid. It is consistent within a certain range of values. For example, if sizes go from S-XL a valid value cannot be XXXL.
• Integrity. It contains coherent values and proper relationships.
• Reasonable. The data is reasonable. For example, it doesn't say someones income is negative or $30 trillion.
• Unique. The data doesn't have duplicate values that are exact matches. Some records may have common features such as 2 John Smith's but they can still be unique.
• Current. Is the record up to date or is it dated?

How To Get Quality Data?

The best way to get high quality data is you are Highly Intentional about it. You will likely want to do at least some of the following.

• Standards. Your organization has standards for data entry that are written, specific and measurable.
• Processes. You have quality control processes that prevent poor quality data from being used.
• Automated Checking. You have automated systems that check data for quality and notify you when something is off.
• You remove silos through integration. Data is highly relational. A silo is when data from an app is stuck there and can't be used outside of that app. You need integration processes.
• Appropriate access. Not everyone needs to be able to access all data but everyone should be able to access the right data to do their job well and make good decisions. You need policies that set this.
• Easy to use systems. Most people are not a super techie like me. You needs systems that deliver quality data to the right people when they need it that are easy to use.