What If we're overthinking Data Products?

Does a Clear Definition Really Lead to Better Outcomes?

Be sure to subscribe if you have not already!

The confusion in definitions

Every time I speak to someone about data products, the topic always comes up. “What’s your definition of a data product”? Even speaking with

Joe Reis

recently, he shared that at Data Day Texas this year they had a room full of data practitioners discussing the topic, but still didn’t come to a conclusion, or at least a conclusion everyone agreed on. Is the data industry alone in all this definition wrangling?

I’ve attended executive lunches where they have defined data products as reports, and I’ve spoken to practitioners who proceed to define data products by breaking down the mandatory illities a data product must have.

So why do we worry about defining things so much, does it really matter? I recall the data industry struggling to define semantic layers, and a quick search it seems we still are. Do other parallel industries squabble about definitions so much? What about software?

Moving to a semantic layer” but I have no idea what they actually mean and at this point I’m too afraid to ask lol ¹

What about software or SaaS?

The term software has been thrown around since the 1958² and, believe it or not, they also debated on its definition and meaning. Decades later, it’s in the dictionary and no one really cares about the definition (at least I don’t). I even had to look up the official definition. Do you care about the definition?

Software: the instructions that control what a computer does; computer programs - Software

Software is what tells hardware what to do. Simple enough right?

• Soft: non-physical, flexible nature of software, it can be changed, copied, or updated, unlike physical hardware.

• Ware: A type of product, in the case of software, a digital product.

So, software is a product that tells hardware what to do and there are a lot of different types of software, e.g. firmware, malware, ransomware, vaporware. This sounds exactly the same as a data product, or at least close enough for most people.

So why didn’t we call it Dataware?

I am not attempting to introduce a new square to buzzword bingo with the introduction of “Dataware”, or do wordplay, but it begs the question on why are we so caught up on defining it when a data product is just a flexible, non-physical product that enables the efficient access, and use of data across systems and users. Should we just start calling it Dataware? No, let’s not confuse the industry more than it already is.

There are only two hard things in Computer Science: cache invalidation and naming things - Phil Karlton ³

You may argue! Yeh Ash, it is different. Data Products have a strong product focus, and they are treated like products. You are right, but so is software and SaaS and I’m sparking some discussions.

Wrapping up

This post was a bit of a reflection post if anything, I’ve been thinking about defining and creating clarity on the definition behind data products a lot for my book and have come to the conclusion it’s not really a big deal and eventually data will have its SaaS moment, and we’ll all just focus on delivering great data products that deliver value.

So, to ring true to the title of this article, I’m not going to give you my definition of data products (in this post at least), and you’ll never hear me use the word Dataware.

What I will close out with is my suggestion to not get caught up defining data products. Data products are, at the end of the day, just software with data as its core value driver. It’s important to have a shared definition within your organization and the industry so we can all create alignment, reduce ambiguity, and better collaborate.

What do you think?

Thanks for reading!
-Ash

Extra reads

• Data Products, Interoperability, Data Teams, and More -

  Joe Reis
  /
  Ash Smith

• https://www.nextdata.com/our-pov/introducing-nextdata-os-autonomous-data-products-for-the-autonomous-era

• https://simonwillison.net/2025/May/31/snitchbench-with-llm/ -

  Simon Willison

1. https://www.reddit.com/r/dataengineering/comments/1e8bn9v/comment/le66un1/

2. https://en.wikipedia.org/wiki/Software

3. https://skeptics.stackexchange.com/questions/19836/has-phil-karlton-ever-said-there-are-only-two-hard-things-in-computer-science