Thoughts from a Talent Analytics leader

Data scientists (and other technical analysts) tend to be mathematically & statistically trained, confident in our numerical and programming skillsets and our abilities to solve problems. 
Sounds great, yes, but is that enough? 

Pete Kropman, at the time Head of Quant Talent Analytics for a bank, shared his views on the 'EXPLORE - The Future' podcast, discussing (among other topics): 

• typical gaps technical people have, 
• the importance of time and learning by doing, and 
• the extra skills managers need. 

While real business problems benefit a lot from approaching them with a data-driven and scientific mindset, other things are important too: 

1. EQ and communication skills
2. business understanding and a commercial perspective - domain knowledge helps 
3. an execution and implementation mindset 

Pete says that these are typically not well developed when a new data scientist starts, and require deliberate attention. "It's great to be able to access all those Python libraries and build an advanced model with a few lines of code, it's another thing to really understand what the drivers of profitability in a certain business unit are, and how to connect the dots between [our] technical work and what's actually going to make a difference to driving the business forward." He warns against building sophisticated predictive models that aren't implemented realistically for the needs of operational people, sales people etc. Especially in a highly-regulated environment like a large bank. Instead, he recommends focusing on "learn[ing a] mindset of bias to execution and delivery". Understanding our users' needs is particularly important i.e. empathy. 

Beware of disillusionment setting in though (as we realise we won't be spending every day building models and generating insights). Pete points to realities like:

• sourcing and cleaning data, 
• learning how to communicate our work really well, 
• navigating approval committees and politics, 
• ironing out deployment issues with other departments, and more...

There are no real shortcuts: it will be a fairly long and sometimes frustrating process to push through to really understanding the business and learning how to get stuff done. Practice is required. We should prepare ourselves for the learning struggle, and call on others for help. Tough yes, but doable - others have overcome similar challenges before!

And, when we manage others, there are additional skills needed there: motivating our team, inspiring them, setting direction, managing priorities, managing workflow, and syncing between different stakeholders are examples Pete gives. Management and leadership are things we can't learn simply from reading a book or attending a course - generally we only get better by doing it i.e. trying and failing and learning by doing. 

Thank you Pete for sharing your experiences of leading analytical talent, and for forewarning us that it won't all be plain sailing. Reminds me a little of Ray Dalio's mantra to 'embrace reality'. 

A lifetime of learning lies ahead - isn't that exciting?