Data Modeling 101: Entities

Tucker Fischer
3 min readNov 10, 2024

--

When we talk to each other there are a million things left unspoken because of our assumptions, shared understandings, and presuppositions.

Indeed if you’ve learned a new language you know this for fact. Speaking a different language from your mother tongue isn’t just rote translation. It’s communicating ideas, feeling, and concepts.

In the business world, we have to be more careful about communication. Often times people in the same company don’t have a common background and shared understanding.

Many businesses call different things the same and the same things by different names. In large enough companies there’s even variance in the meaning of words, and ideas depending on what business unit you are in.

Take the classic example of parts and suppliers. A part could be a single thing, or it could be a conglomerate of little parts. Often the definition depends on what the subject person cares about.

Parts and suppliers are both considered entities which is the subject of the rest of this article.

Entities in the broadest sense of the word just means a thing.

A part is a thing.

A supplier is a thing.

A person is a thing (Although don’t say this normally as it would be considered rude).

An order is a thing.

So on and so forth.

Things matter because without things you cant have things.

So maybe using the word thing isn’t a great way to describe what an entity is.

“An entity is a thing of significance about which the organization wishes to hold (which is to say, collect, maintain, and use) information.” — David C. Hay

The above definition is the most common that you will come across, and has been defined as such in one way or another for the past 50 years.

At face value an entity is a pretty easy concept to understand.

When the rubber meets the road, in the real world, though, it’s not always cut and dry.

Take for example a company that fulfills eCommerce orders and ships the orders to the end consumer. In this context the “shipment” entity is of vital importance.

But what constitutes a shipment?

Is it any package that leaves the warehouse? Or is it any package that has a tracking number from a carrier? Is it any one package that has a tracking number from a carrier and leaves the warehouse?

In his book ‘Data & Reality’ William Kent brings up three things you need to think about when trying to draw the line of what is an entity.

Oneness: ‘What is one thing?’

What is one shipment?

Sameness: ‘When do we say two things are the same, or the same thing?’

Are Direct-to-Consumer shipments the same as Less-than-Truckload shipments? Are they the same as Truckload shipments?

Categories: ‘What is it? In what categories do we perceive the thing to be in?’

What do we (the business) mean when we say shipments?

Your answers will likely differ from others.

What’s important isn’t necessarily to have the same answer as other businesses but rather that your organization has the same answers.

Because entities are the bullwork of your organization. They help your organization operate with the same understanding, and context which ultimately reduces friction in communication, leads to better, and quicker decisions as well as countless other benefits.

--

--

Tucker Fischer
Tucker Fischer

Written by Tucker Fischer

0 Followers

Data management consultant helping eCommerce fulfillment companies build scalable, efficient data systems.

No responses yet