Semiotics = Branding
Let’s start with a few definitions and then we’ll take off running.
Semiotics, the discipline, concerns itself with how we attribute, create and use signs.
A sign is anything that combines a signifier and a signified. For example, “bachelor”. The signifier is the unique composition of letters. The signified is the concept/meaning: an unmarried male.
You might be thinking that it’s a bit silly to have a discipline that studies something we do automatically and unconsciously.
Or you might be thinking, “But what about those signs that don’t have a one-to-one relationship? The word “flag”, for example, does it mean a piece of material or the verb to waive to get someone’s attention?”
And once you’re locked into the second train of thought, you find it’s a wild ride. Concepts such as justice, good, bad, beautiful… all jumble up into mutable conditions with countless exceptions, and you wonder how we communicate at all.
And, if you’re still reading this, you probably understand how to create, detangle and reinforce signs, or, in marketing speak…
You understand how to create a brand.
And you understand that branding is much easier if you’re able to interact with the audience. You most likely appreciate focus groups and feedback. You probably pay close attention to social media and/or analytics.
All in all, you pay attention to how well you’ve matched the signifier to the signified. Did your intended meaning come across? Did the audience understand, feel, intuit, what you were trying to communicate? Have you been specific enough? Too specific? What elements (colors, copy, images..) are working, which aren’t? How do you know?
If any of these questions keep you up at night, then I’ve got just the thing for you: machine learning.
Machine learning is really just pattern recognition done at a humanly impossible scale.
The computer runs every single scenario and tests its predictive powers.
The best part is that you can add machine learning to digital communication (websites, digital campaigns, social media) to find these patterns. Then, you can add in user behavior and, WHAMO, you have highly correlated relationships among specific branding elements and user behavior.
You’re communicating with your audience… at scale.
Take these patterns, leverage them, exploit them, run the machine learning again, rinse and repeat.
And all those anxious questions start to dissipate. You know if the signifier you delivered the meaning you intended. You have data to prove it.
Thank you, Machine Learning.
Author: Randall Malcolm