Let's start with some Data Science definitions

What would your answer be to the question: "what is data science?"

Hi dream team,

Let’s add a bit of definition to what I was thinking

So - what would your answer be to the question: "what is data science?"

Well let’s get down to it. See below an image I reach for when I’m asked this question, since I like it the most and feel the overlaps are broadly representative of what is going on.

First let’s get it out of the way and do the left hand side. So Artificial Intelligence (AI) is this broad area of having computers be able to make some kind of decision by looking at some training data, then being able to decide something.

AI is the umbrella term for “everything” but now and in the past. In the past it was more a general collection of techniques occasionally referenced as “AI” - now every single thing and widget in the world is AI.

Machine learning (ML) is a subset of AI. ML is limited by the fact that algorithms make predictions or understand structures in data, but the limit is ML only knows what is in the training data. Working backwards up to AI for a second - what you know as Generative AI does have the ability to create new data based on a triangulation of knowledge contained in a large language model. Hence it is differentiation from ML - and this “creation of new information” is what was so exciting when these language models hit the mainstream.

Deep learning (DL) is a subset of ML and is a special category of models that very complex and well suited to very complex data structures - such as working with image data (it is a cat or a dog, is it a ball bouncing in front of this Tesla, etc). Its also a type of model that’s particularly unsuited to marketing data. I’ll write about that in future. But for now if you hear “Deep Learning” solutions that are gonna help you crush it in your marketing life, its a red flag.

So what about Data Science? Well it cuts over a few of these other categories. I may present some musings on AI but its not at the heart of what I want to do here, there is a tonne of information on AI swirling around. ML I’ll 100% be getting into. DL less likely but I will explain my comment above so you have an opinion on that.

But what about that big zone to the right?

So the overlap between Data Science and those other categories means Data Science gets to dabble in any and all of those things (AI, ML, DL), if the need arises. You don’t have to have a PhD in AI or be a propeller-head to do well with Data Science btw - in my opinion the biggest ROI is curiosity.

Back to the big zone on the right - well what’s in there then. I’d say it was these things:

1. Analytics, which probably captures both:

   1. KPIs & Metrics (AKA “reporting”)

   2. Legit analytics (I will explain another time what I think characterises this)

2. Working with others / capturing hearts and minds, which includes:

   1. Presentation skills which is often described now as “Data Storytelling”

   2. Collaboration skills - those tactical touches and people skills that means you get the input you need to do a great job

3. Business knowledge, which probably manifests in a couple of ways:

   1. General business literacy - helps to know the goal right?!

   2. Consultancy skills - is to me, the alchemy of being able to “listen and explain” -

So what does all this mean?

Well, hopefully I’ve painted the picture here that Data Science is a technical topic where people work with data in a similar way to an AI researcher, or ML developer. But the clear space that is unoccupied on the Data Science bubble is these other skills that help you gets things done. So there’s “what you do” and “how you do it”.

My plan for Data Science for Marketers to demystify, explain and describe how these technical things help us with marketing, but in equal measure talk about:

1. Provide a perspective on all the analytics that happens in marketing, so you know what to ask for, and when.

2. How to engage people with data so you can do the same - tips for building data/data strategy decks so you can crush your meetings.

3. Shed some light on how to have more fruitful relationship with your Data Science people.

4. Using data to inspire change, reduce uncertainty etc.

5. Practical tips that will help you understand key analytics concepts.

6. I’ll share some practical things I’ve been doing for 20 years whilst working hands-on with data.

So you can expect a strong focus on the 6 things above (maybe I forgot some and there’s more) - what I don’t expect is for anyone to run off and start learning to code as a result of reading anything here. What is my big hope though is I can help with a bit of an upgrade in general knowledge around Data Science, and how it helps us here in marketing.

I will likely record / share some video content to make things interesting as we go along. Looking forward to getting into into it all. Thank you dream team.

Cheers,

John