Issue #3 - Learning Curve, A Strategic Tool and my Data Science Toolkit
Hey all,
I hope everyone is keeping well.
Over in Ireland, it’s starting to get a little cold, -1°C this morning, and I’ve been sitting in my little home office wondering why the room is so cold - only to find out I had forgotten to switch the heating on 🤣
Welcome to Issue 3, I wanted to share 3 topics this week:
Four Corners Analysis Tool - I find this tool to be one of the most formative tools I have used. If done correctly you should be able to understand people, teams and organisations in greater detail.
Learning Curve - You’ve probably heard of it before, maybe seen it, but I wanted to share a more holistic picture - maybe it will help you think about your own growth.
Data Science Tools - for most, data science is cool and interesting. And many businesses want to become more ‘data driven’ - well I am a data scientist and I wanted to share a list of the tools that I use. I kept the list short for this week, but I’ll expand on the list over time if people are interested.
Thanks for opening this, wishing everyone a happy Friday and a great weekend.
MBA Bitesize - 4 corners analysis
Porters 4 Corners Analysis - it will help you to try and understand what other people, teams and organisations will do.
While it’s aimed at ‘competitors’ I would encourage it to be applied to situations where you want to understand both the causes of behaviour and also the action they are likely to do.
I’ve applied this many times, one example was around team planning where we had some overlap with partner teams. In order to work more effectively with those teams and to also develop our own team’s strategy I really wanted to consider the situation from their perspective:
Drivers - what are they trying to achieve? what are their goals?
Assumptions - what do those teams believe to be true?
Current Strategy - how are they executing their work?
Capabilities - what can they actually do? what are they really good at? what are they not so good at?
By taking these four elements/corners into account I was able to better understand my partner teams, understand their likely future strategy and develop my own team’s strategy.
Here is a link to read more.
Personal Growth - Learning Curve
We’ve all probably heard of the learning curve, perhaps we’ve even said something like “it had a steep learning curve but I got there in the end” when learning a new skill.
A typical example we can likely share is when starting a new job. It takes a while to learn how to do the job before we know what we are doing, i.e. the learning phase plateau’s out and we stop learning so much. The example can be seen in the diagram above (black line).
Some examples from my career. When I was 18 I was flipping burgers at Mcdonald’s, it took about 1 hour to be taught that. I left Mcdonald’s to work at BAE SYSTEM to build military aeroplanes, it took 3 years to learn how to do that job. In both examples I had to learn a new skill, one was quicker and less intense than the other.
But what happens once we’ve learned that new thing?
Simply put - we mostly keep learning - just not as much. For example - while I was learning to flip burgers, I was also learning about the company culture, working in a new team, health & safety, plus a whole bunch of other stuff. This was the steep stage of the learning curve. When I moved on to learning how to work on the drive-thru I only had to learn that part of the job, therefore the amount of growth was a lot less steep.
The key difference between learning is 2 parts, firstly the cadence in which new learning events happen, and secondly, the growth opportunity/size of the learning.
Look at the black line. An example of this person could be someone who picked a career/job and did it for a very long time. They joined a company when they were 18 years old, learned the job early in their journey and stayed there for 47 years before retiring at 65. I’ve not met anyone who has done that for 47 years, but I have known people who have done that for maybe 20 years.
So the red line - this is someone who was maybe hoping to follow the black line but some event happened (exclamation mark). Perhaps their role was made redundant, perhaps a mid-life crisis, illness, etc. At this point, they have maybe been doing the same job for a good amount of time and their growth kinda plateaued out. I’ve seen many examples of this, for example, people I worked with a BAE SYSTEMS, huge redundancies and people leaving their roles after 10 to 20 years now forced into a change.
My career looks more like the blue line. The events are all different including “I didn’t want to flip burgers all my life”, “I wanted to move from building aeroplanes to working in computer science”, “I wanted to teach and inspire people”, “wow this data science stuff looks cool, I’d like to do that”, “I think I’d be good as a people manager”, etc.
Growth is generally seen as a positive thing. And sometimes it’s easy to deprioritise it (black line) and then those unexpected life-changing events can ‘force’ a change. Alternatively, you can take ownership of that growth and it can come in many forms - a short online course, changing roles, asking your manager for a little more responsibility or starting a personal project.
So what’s the point of this section? I guess to get you thinking. Ask yourself, which line are you on right now (Black, Blue, Red)? Can you see any ‘events’ coming up? Do you want to pre-empt those events and start something now? Or just see what happens when you get there? There isn’t a correct answer and each of us is different.
My day-to-day Data Science Tool (Part 1)
I know a lot of people are interested in data science (DS), statistics and machine learning, so I wanted to show the tools I use on an almost daily basis.
Note: I’ve split this into multiple parts - mainly because it would be a huge section.
Let’s just get straight in…
Python - it’s my go-to language for anything DS related. Another great alternative is R. I used R during the first year of my PhD and then went on to teach it with the amazing Magnus Johnson at the University of Hull. I liked R, but I looked at the jobs I wanted to go into after my PhD and saw most were asking for Python. So I switched. Today, I see more people lean towards Python for DS but people in my team use R too.
SQL - hopefully, goes without saying, SQL is super important. I met a person once, amazing at statistics, and amazing at finding data-driven insights which pushed the business forward - however, didn’t know SQL and therefore was reliant on other people to get data for them. In my experience, a lot of business/data problems can be answered at the data extraction stage with just SQL. As we start to think about big data, Presto becomes more common - syntax is very similar to SQL so it’s not too much of a jump. My tip - focus on MySQL queries, it’ll cover everything you need.
Jupyter Notebooks - I use this for nearly everything. I’ve used R-Studio in the past (for R obviously) and occasionally use VS Code if I’m turning any DS projects into anything meatier. This could include creating data pipelines or even creating APIs or standalone projects.
Pandas and Numpy - don’t hate me, I put them both in the same line because I use them interchangeably. This is your core way of handling data in Python for DS purposes. Both have very useful functions such as basic statistics stuff such as mean, standard deviation, etc and also some basic plotting functions.
Matplotlib - this is my main visualisation library. Seaborn and Plotly look nicer out of the box (which I still use), but Matplotlib has become my go-to tool for visualisation.
In the interest of brevity, I’ll leave this list there for today and I’ll talk about some other tools next time. In all honesty, you can probably solve most data problems with just those tools.
If you’re interested in DS, these are a good starting point.