Perspective 6: More About Data (Part 2)

In my last blog post we started discussing the topic of data. We discussed how to identify data, and one of the tools (Business Glossary) you can leverage to ensure everyone is speaking the same language.

In this blog post I will take you through the 3 other techniques that may be leveraged to document the data as well as a few helpful hints on how to best analyze data. To be clear, these are not the only techniques, just the last 3 of the 4 I wanted to discuss so let’s jump in.


The purpose of the data dictionary is to identify and define the:

  • Data Element – a specific value that has a specific meaning.
  • Description – the defined meaning of the element.
  • Data Type – the category of data defined by the value(s) it can take.
  • Length – Specific requirements on how long or short an element name can be.
  • Required – Defines if the element is required or not
  • Default Values – Defines if the data element needs to have a default value or not. The default value is a preexisting value.

There may be other attributes that can be captured based on the organizational needs, but these are a great started.

So, building upon our example from Part 1 for interacting with a financial institution, you may have data elements defined as below that relate back to specific data that would be captured.

Data dictionary

Figure 1


An entity relationship diagram describes entity types and specifies the relationships that can exist between them.

Again, building upon our example from Part 1, three entities that may exist in the financial data structures are “Account”, “Customer”, and “Product”.  In Figure 2 we see that within the “Account” entity you may have data elements such as “Account ID”, “Account Name”, “Account Description” and “Account Open Date”. Within the “Customer” entity you may have data elements such as “ID”, “Name”, “Address”, “Phone”, “Email”, “Account Number” and Type”. Now the relation between the “Account” entity and the “Customer” entity for this example is as follows:

  • An Account could have 1 to many Customers.
  • A Customer could have 1 to many Accounts.

To be considered a customer you must have an account with the financial institution. For this example and account and product are different. The account is either a checking or savings account. A product may be a credit card or personal loan.

Customer mapping

Figure 2


A system context diagram demonstrates the external components that may interact with the system. I like to define a core system and then show the other systems, applications, or other external components that interact with the core system. In Figure 3 we see the core system named the “System of Record” and the other systems or application that interact with that core system, as well as the data that is passed between those systems.

System context diagram

Figure 3

Now that we have completed reviewing the last 3 techniques, I wanted to focus on let us move into how to analyze data.

How to Analyze Data

There are many ways to slice and dice data, and it can be fun doing so. I like to pursue analyzing data in 6 steps:

  1. Step 1: Determine what type of data you need to answer specific questions to solve problems. Make sure you level set on the purpose of obtaining the data and the desired outcome. When diving into ambiguous and unstructured data, you should define hypotheses to validate though the process.
  2. Step 2: Collect data depending on the requirements you defined in step 1. This data can come from a wide array of sources like we discussed. You can get it through reports, databases, data warehouses, processes, customer satisfaction surveys and so much more. You can also conduct your own focus groups, interviews and even observe how individuals do their work.
  3. Step 3: Scrub the data. With an initial data set, you may find missing, incomplete, or repetitive data, which can bias the results. You will want to check for outliers and ensure metrics, like the mean, median, mode, and range, make sense given the context. Sometimes you also need to convert data into a format that is readable by data analytics tools.
  4. Step 4: Analyze the data through leveraging tools in your organization. Make sure your analysis goes back to the objective you want to achieve conducting the data analytics. Look for patterns, significant variances, and data that looks out a place and does not fit with the data set. Start asking questions for items that don’t make sense to you.
  5. Step 5: Compile the results. Once the data is collected and analyzed, compile it, and organize it.
  6. Step 6: Present – Present it in an easy-to-understand format. Many companies have internal dashboards that track Key Performance Indicators (KPIs) through graphs and charts. But if not, know your audience and know the story you want to tell with the data. For example, if you are using the data to help drive a business case for a project that will render huge transformational change for the organization, you want to make sure the data aligns to that objective.

Final Thoughts

In my next and last post on this topic, I will give a brief introduction to data analytics, including the various types and how to leverage various techniques to analyze data.

Be sure to connect with me on LinkedIn or follow me on Twitter.

Until next time, signing off,
Paula Bell
The BA Martial Artist ????

Related Content


Creating a Digital Twin of an Organization: Three Steps to Success 

By Max Smith, Senior Director of Product Marketing  Buzzwords come and go in the business world, but one of the most persis...

What is a Process Repository & Why Does it Matter? 

By Dylan Boynton,  A process repository is an essential tool for any organization looking to optimize their processes and a...

From Business Process Diagrams to Business Process Models: What Every Company Needs to Know

By Max Smith, Senior Director, Product Marketing For many companies, the journey toward business process excellence starts w...