The importance of semantics

Something I’ve become quite self conscious of in recent years is that I “talk funny”. Working in consultancy and having overseas clients, I’ve become aware that I often use various turns of phrase that aren’t quite as international as I might have thought.
One of my most memorable occurrences of this was using the phrase “teaching granny to suck eggs” in a meeting with some US based clients, which was met by the people on the call with quite a lot of confusion until I explained myself.
Conscious of the fact that this publication does have a multi-national audience (thanks for following by the way!), I feel the need to include a pre-amble on today’s subject to set the scene.
The Great British Bread Debate
You may have heard of the claim that Eskimos have over 100 different words for snow. Well I reckon this statistic is surpassed by the number of different words an Englishman has for a bread bun.
For a relatively small island, it’s a lot. The following infographic I’m fairly sure doesn’t cover them all:

As a life long northerner, I’m fairly sure “bun” was my choice of term growing up, but having relocated to Lancashire in my 20s, I’ve since adopted “barm” as my preferred nomenclature.
If you’re ever feeling mischievous or want an opportunity to sow some division amongst a group of Brits simply ask, “what’s the right word for a bread bun?” and based on region and upbringing you can guarantee it will spark a heated debate.
Quite why this bun-fight has endured so long (pun intended) I’m not sure. In fact could this particular disagreement be the origin of the term bun-fight? (and if I’m making puns about bun-fights is it a pun-fight?!).
Anyways, point of clarification number one: The British have many different words for a bread bun.
Greggs the Bakers
The second piece of cultural nuance I feel I need to introduce is Greggs the Bakers. Greggs is a chain of bakeries here in the UK. Their most famous (notorious?) product is the sausage roll, a baked good that consists of sausage meat wrapped in puff pastry. I don’t think Greggs invented the sausage roll. They are just renowned for making them, to the point that their brand is almost synonymous with the flaky meaty treat.
If you’re not familiar, this video clip shows Hollywood star Ryan Reynolds sampling this British delicacy:
Craving Breakfast
So my train of thought here is spawned completely from feeling hungry earlier this week. I’m not normally a breakfast guy. A cup of strong coffee is all I usually need to get me through the morning to lunchtime. But for some reason, I woke the other day famished. So on the walk into the co-working space I try to frequent a few times a week, I figured I’d treat myself to some breakfast from a Greggs.
I had a real craving for a sausage butty (another colloquialism right there for you… butty is a slang term for a sandwich).
When is a Sausage Roll not a Sausage Roll?
If you take a look at the Greggs breakfast menu, have a guess which term they have decided on to describe their bread buns?
That’s right: roll
WHY!?!
Greggs was actually founded in Newcastle-Upon-Tyne, and is a brand I’d would commonly associate with the North. “Roll” as a regional term has always been used more down South (as an aside, where the North and South of England start and end is another fiercely contested battle if you fancy winding up your English friends even more).
But if you haven’t figured it out already, this makes even less sense when you walk into a Greggs and order a sausage roll. Thus begins a Russian roulette-esque game of the fates (sausage roll-ette perhaps?). Will you receive the breakfast barm of your dreams? Or will it be a warm and comforting flaky snack as a consolation?
On this occasion, the gods were smiling on me and I was served with the porky produce I had intended (though I was later alarmed to discover they’d given me tomato ketchup as opposed to the requested HP Sauce).
Words are important
The lesson to learn here for data professionals is that words are important. If you’re building a; data platform, a data product, a report, a dashboard it’s important to give things the right names and make sure you leave no room for ambiguity.
Heck, even when writing a simple SQL statement ambiguity can easily bite you in the ass. I dare you to deny that you have never seen some variant on the error message below:

Semantics, the meaning we attach to words, really do matter in the data world. You may have two stakeholders from two different domains who use the word “revenue” in different contexts. For example, a sales manager might be interested in the sales made by a sales person on a given day. This figure is based on the day of transaction and doesn’t take into account things like cancellations after the event. Where as a Finance manager will probably be more interested in the money that actually hits the bank.
Rather then re-use the term “Revenue” in two different places, data professionals should strive to be more deliberate with naming, adding context and additional meaning to data terms to help save the confusion of whose definition of a particular meaning may be correct. This can then help save bun-fights in the boardroom!
In the real-life example above, we ended up with a “Sales-In Revenue” metric and a “Recognised Revenue” metric.
With Enterprise Data catalogues and features available in metadata management catalogues, the ability to provide further context can be extended even further, making sure folks provide relevant descriptions for data fields and or metrics.
With my Databricks hat on, make sure you’re using Unity Catalog to describe your tables and columns. And the Power BI nerd in me would also encourage you to use the descriptions fields for your Semantic Model measures, tables and columns. Data modelling tools such as Ellie allow you to add this level at context as early as the conceptual modelling stage.
Communicating Clearly
Being clear in your naming and leaving ambiguity at the door can only help with making data more usable and intuitive for end users and ultimately being able to clearly communicate with data is a key driver for companies that wish to succeed.
One of my favourite quotes on communicating with data is from Stephanie Evergreen:
As I’ve touched on in a previous article, adding all this additional context serves an added purpose as not only does it make life easier for humans, it makes life easier for machines, and well thought out semantics will only help in the AI age.

Conclusion
So, words matter. How you use language when presenting data is vital for comprehension and understanding, so make sure you give it some thought and are deliberate with it.
Now, all this typing has made me hungry. I think I’m going to go and eat a sausage roll.
2 Comments
Antony · February 7, 2025 at 12:06 pm
Great article Johnny. I get on my high horses when users say dashboard when they mean report in a Power BI context
Bubba · February 7, 2025 at 12:18 pm
REALLY enjoyed reading this. And you managed the entire piece without employing the “V” word!
(How can you discover whether or not someone is vegan? Don’t worry they WILL tell you!)
The older (and grumpier) I get, the more I realise the need for absolute clarity in communication.
Great piece!