There is a misconception that people working in the field of Data, Analytics, and AI are academic types. This conjures up images of rarified enclaves of computer science minds, debating how to design the optimal model to crack some arbitrary performance benchmark. While it is true that data scientists can approach questions from quite a theoretical slant (see Kaggle), I have found that data practitioners tend to primarily care about real-world impact.
Yet having impact with data – being a data leader in the broadest sense – is not easy. As an article in the Harvard Business Review asked last month: why is it so hard to become a data driven company? On a personal level this is not a straightforward question either. While I have written about how teenagers should pick data as a career, I sidestepped the question of what specific role to choose. If you want to really move the needle with data, where do you go?
Should you focus on developing technical expertise? Where do soft skills come into this? Perhaps you could better prioritise opportunities that give you access to interesting data? How about business or industry exposure? Like many people, my career began with developing my technical skills. While for close to a decade I was convinced this was a surefire means of maximising my impact with data, more recent roles have underscored that this is usually not the case.
More than fifteen years ago, I kicked off my journey with programming skills (hello PHP 😂) as I had read that this was a good place to start. I quickly learned I needed to brush up on SQL in order to make sense of data. As I started building products with real users, I moved on to data visualisation. Then in 2010, as I became more involved with social media platforms, I realised I did not quite know how to approach data in this domain.
To learn more about the intersection of data and social science, I decided to pursue a master's degree at the Oxford Internet Institute. I learned how to navigate data at scale, using tools such as social network analysis. However, I quickly encountered my next problem: access to good data – something difficult to achieve on a student budget. Starting a role at Google therefore was a logical next step; all the data I could ever want, from Search to YouTube.
Over the next years, I had the opportunity to work in a range of different roles akin to that of insights analyst, data scientist, transformation lead, and data translator. While highly enjoyable, I ended up moving to Dyson because I realised that Google was a bubble. As clients used to tell me: "Your insights and analyses are spot on, but you recommendations are not – you simply do not know what it is like to be in our shoes." Fair point. Time for some new shoes.
Arriving at Dyson, a traditional engineering company with an iconoclastic culture, quickly taught me what those former clients had been trying to tell me: It does not matter how clever your strategy or neat your technical solutions, without the right talent in place and engaged stakeholders, little will happen. Not surprisingly, a significant part of my time there went into recruitment, training, and establishing an analytics community across the organisation.
While I am not discounting the value of my technical expertise, Dyson drove home the point that analytical capabilities are usually not the limiting factor in delivering impact as a data leader. Crucially, there needs to strong senior sponsorship. You need an executive committee and board that are fully bought in for a data transformation to be truly sustainable. Luckily, this is exactly what I encountered when I was approached by the Chalhoub Group a few years later.
The last eighteen months, although challenging, have demonstrated what can be done when the stars align on budgets, executive buy-in, talent, stakeholders, etc. And yet, there are still so many opportunities to seize. Although our data team is executing as well as could be asked of anyone, how can we maximise our impact? From the technical to the managerial, there are so many data roles to choose from. In the long run, what role should we as data leaders play?