The Power of Definitions

The line looks way too long as I and many other weary red night warriors approach the end of the customs and immigration queue at Heathrow Airport in London after the overnight flight from Newark. Sometimes the wait to get through and collect my luggage takes less than 20 minutes. However, this morning it will probably take at least two hours. Step by step, we work our way through the queue as it snakes back and forth until finally, I am handing my passport to the immigration official.?

After answering her questions about what brought me to the UK, she filled out a short form and asked me to wait on a bench until called. This has never happened before, and I tried explaining this to her. “I teach a data modeling class here in London twice a year, and I have been doing this for many years.” She explains that I need a work visa to work in the UK and points again to the bench where I need to wait.?

After about an hour on the bench, I heard my name and followed somebody to a different room. This room contained a few other people, all waiting as I was now to decide whether I could enter the country. The room contained an old television, a phone, and a few wood benches.?During the next seven hours, I was interviewed by a series of immigration officials, each asking similar questions about why I was here, and each ending with that I needed a work visa. This was not going well. However, during the last interview, I decided to use my data modeling superpowers. The woman started asking me the same questions about why I am here, but instead, I countered by asking some of my own questions. Definition questions. The conversation was similar to this.?

“Why are you here?”?

“I am teaching a data modeling class here in London. (I then turned on my data modeling powers.) Under what conditions would somebody not need a work visa? After all, I have taught here many times before.”?

“If you work for a US-based company and show me papers explaining your purpose.”?

“Ah, I do work for a US company…”?

“Mr. Hoberman, your name is Steve Hoberman and your company is Steve Hoberman. So you are one of the same.”?

“How does the immigration department here define an Individual? How do they define an Organization?”?

This was the turning point in the conversation, and I knew, based on her hesitations, that this agency lacked a very clear definition of these two terms, and within two hours, I was unpacking in my London hotel room (and still trying to catch my breath).

We data modelers ask lots of questions to ensure the business picture is precise. We build diagrams with business terms, properties, and relationships that tell a story without ambiguity so that all readers of the model read the story the same exact way.?

But what about the definitions? How do you define an individual? An organization? A customer? A product??

Definitions must be equally precise. I spend quite a bit of my consulting hours reviewing data models (Data Model Scorecard ), and I very rarely read a good definition. I see a lot of missing definitions, a lot of useless definitions (“Order date is the date of the order.”), and a lot of ambiguous definitions.?

If you’ve taken my data modeling class (our next Data Modeling Master Class is in September), you know that we spend time discussing the Three C’s of definitions: Clarity, Completeness, and Correctness.?

Let’ take about just one aspect of clarity on writing a good definition, and that is the lack of ambiguous phrases, such as these:?

• In general?
• Sometimes?
• Most of the time?
• Normally?
• With new exception?
• Often
• Maybe?
• Almost always?

These phrases need to be removed or explained to make the definition precise.?For example, here is a partial definition of Customer with an ambiguous phrase:?

A Customer is a person or organization who obtains our product for resale. The Customer normally obtains the product through purchase.??

We can either remove the word “normally”, or explain it. Here it is explained:?

An example of a customer who does not purchase our product is the Salvation Army,?which receives the product for free as a charity organization.??

I can talk for hours about writing good definitions. But instead, like a good Netflix series, I will keep you in suspense until the next edition. Until then!?