Make it easy as π

Finn Shewell
7 min readMar 9, 2021



No matter what you do, you’re good at something.

As you get better at something, you tend to focus on it more — we all do, creating a virtuous cycle that ends in everyone becoming experts in some niche domain.
If you’re interested in UX now — for example — 10 years from now you could well be the best in the world at funneling traffic to conversion through smart web design. If you’re really digging the original timber floorboards at your new place, maybe your next instagram account will be about refinishing hardwood floors at a world-class level.

So we’re all growing these potentially world-class skills, all the time. Let’s talk about that — and how we can make that happen in a new way.

The basic form of expertise I’ve just presented is called an ‘I-type’ skill set. Based on Ideo Founder Tim Brown’s ‘T-shaped people’, the I-type skill set is born from a deep engagement in any given area. An I-type skill set is defined less by what it is, and more by what it isn’t. I-type individuals exclusively have depth of skill — they lack breadth.

With an I-type skillset, someone might be amazing at understanding numbers and thinking quantitatively, but poor at data visualisation, communication, or even understanding what metrics to prioritise. This is because all of these supporting skills require collaboration & empathy — the fundamental ingredient to the horizontal stroke of Tim’s T-type skillset.

Throwing the theory into reality

Now, let’s follow someone (we’ll call them ‘Sam’) as they develop their T-shaped skillset.

Let’s say Sam is at University, and has developed a special interest in marketing. Specifically, how social media can be used to create communities that unite around concepts and causes. Sam’s taken all the relevant papers, and she’s started meeting with industry leaders to learn more about the cutting edge of this discipline. Sam’s developing a great vertical stroke in her T.

Now Sam has graduated, and is working for a rapidly growing Fast-Moving-Consumer-Goods company (Fix & Fogg, anyone?). She’s loving the challenges that come with applying theory to practice, but she’s also hitting some walls. So she reaches out to her team members, and starts to learn more about what they do, how her work feeds into their roles, and vice-versa.
Sam starts to test out some things that will help improve the insights capacity of the organisation and direct the community to a recent campaign they’re running. She also starts to look into some new topics that aren’t quite as related to social media or community marketing — meeting with a psychologist to learn more about the scientific side of what makes communities bond, and taking an online course in analytics to strengthen her quantitative skills.
All of this is really strengthening Sam’s top-of-the-T skillset.

So now Sam has both deep experience in a niche area, and a broad range of skills outside of that area that equip her to better communicate her own findings to her team, and understand their points of view and skillsets in a much deeper way.

Those are some pretty amazing returns for a bit of extra effort — but we can take this another step further. To do so, we need to understand something about learning curves.

Learning curves essentially map out how easy or difficult it is to learn something given your level of expertise in that matter.
We typically perceive learning curves as having either diminishing returns or increasing returns:

When we say something has a steep learning curve, we tend to refer to the curve on the left — large amounts of effort at the start of the journey, with little return. When we say ‘easy to learn, hard to master’, we typically mean the curve on the right — showcasing the ability to progress incredibly rapidly at the onset of your training, but meeting diminishing return as you progress towards expertise.

When we look at learning curves and the T-distribution model together, it becomes clear why we think it’s useful to only develop one skillset in a truly significant way.
Regardless of if the scenario presents diminishing or exponential returns, it either costs too much at the onset to focus on more than one type of skill, or you run into diminishing returns and face heavy headwinds as you approach the leaders and the cutting-edge of that field.
It makes sense, then, to only on one area of expertise, gaining just enough knowledge in other areas to keep up with your colleagues during meetings, and communicate in a way that they can better understand.

But that model is wrong.
Further study has found that we tend to learn along more of an s-curve, where we’re slow to learn both during the first exploratory steps into a topic, and also once we pass a certain level of experience. One could view it as the worst of both worlds.

Let’s take rock-climbing as an illustrative example — at the start, you kinda suck at it. You can’t really hold on to half the holds, you’re tiring out quickly, and it’s tough to see what your next move on the wall should be.

But then, a month later, you’ve become stronger. You’re able to picture your next move more easily, and you understand enough to be able to grasp new concepts immediately and put them into action.

Then we jump forward a year. You’re really good now — you’ve done some V6-V7’s and you’re feeling on top of the world. But you did those V6’s last month, and you haven’t really improved much since. You now know 80% of what there is to know, and your body is as strong as it will get without some serious and regular conditioning work. Your learning is starting to slow.

That doesn’t paint a very cheery picture — sure, the middle sounds great — but if you want to learn any skill, it sounds like you’ll need to put in a lot of effort, enjoy it for a while, then quit once you hit your natural peak.

Let’s change that perspective.

Let them eat Pi

In reality, the s-shaped learning curve is a win-win, and it provides a strong incentive to consciously move away from t-type skill distributions, and toward π-type.

Think of it this way — the initial headwind you’ll suffer when adopting a new skill or area of expertise is something everyone will face, and you’ll see massive drop-off as a result.
If you do any specialised form of physical activity, you see this all the time. People come to your Muay Thai class two weeks in a row, then you never see them again. This barrier to entry ensures only the motivated get through, automatically putting you in a class of your own before you even enter your peak learning phase.

Then, once you start to see diminishing returns — stop. Only learn enough to keep you at your current relative level of expertise in this new skillset — because you can now provide 80–90% of the value, with 50% of the time and resources expended to get to 100%.

Sure, if you want to be the worlds best at eyebrow dancing you could put in your 10,000 hours and make it big — but you can get most of the way there in 5,000 or less.

This new approach to learning alone should be enough to encourage you to develop a secondary super-power — but there are even greater advantages when you start to consider what skills you want to combine.

Let’s say you crush it in the product development space. If you’re looking to create another vertical stroke to your T and make it a π, you have two paths forward. You can choose a highly complementary skill, or a highly contrasted skill.

Complementary vertical strokes have a lot of venn, or overlap. In this product dev. analogy, a complementary skillset might be supply-chain logistics. This is an area of expertise you’re likely to encounter and interact with on a day-to-day basis, and progress in one field directly affects progress in the other. Complementary skillsets offer many cross-transferrable skills, but it’s not likely to differentiate you as much as a contrasting skillset — and the application of one skillset into the other isn’t likely to lead to new insight.

Contrasting skills, however, have very little venn. For the product developer, it could be data science, or IP law. Neither skillset is needed for effective product development — but both can make a product developer massively effective in their role.
There are few transferable skills between the two vertical strokes, but the connections that do exist will provide rare insights that could yield large strides in either field. Very few people are experts in two seemingly disparate skillsets — so you’re in a league of your own, and with the right perspective shift, stand to create unique value in both areas.

So that’s the choice — to pursue a skill with massive transferability at a low-value yield, or pursue one with minimum transferability that stands to produce great insight into either field.

Regardless of the path you travel, you’ve done more than crossed the T’s and dotted the i’s — you’ve made some damn good Pi.



Finn Shewell

👨‍👩‍👦‍👦 I help people work together