We are your data science workshop.

Data Scientist Unicorns Recruitment Guide

Recruiting data scientists can be a big challenge. Often companies are keen to employ good people who know how to make the best use of their data, but don’t know how to choose them. We’ve put down our thoughts into this data scientist unicorns recruitment guide:

What makes recruiting data scientists difficult?

A lot of people are unclear about what it is exactly that data scientists actually do. Certainly, this must make things harder for the recruiter. However, other factors make recruiting harder too:

  • The terminology is unfamiliar.
  • Interview panel members may not know how to test candidates on their knowledge.
  • Not all data scientists are the same. Some have particular strengths in mathematics and statistics, others are software engineering pros, many are expert in analytics and business intelligence. How do the candidates in front of you match your needs?

What can you do to make things easier?

In my opinion, data scientists should be among the best communicators in the tech sector. After all, their job is to take your large, complex data sets and produce easy to understand results. This means they should be able to state clearly what their skills are and how they relate to your needs. But that’s only a start: how can you be confident that you’re making the right hire?

Don’t be intimidated

If a candidate uses unfamiliar terminology, ask them to explain what they mean. Dig down into their knowledge. Some examples might be

  • You’ve mentioned some algorithms. Can you explain to us under what circumstances you would use each of those, and when and why they might need to be avoided?
  • What are the underlying assumptions for those techniques? Are there any mathematics or statistics you could discuss to justify them?
  • Can you discuss the advantages and disadvantages of those packages? In terms of accuracy? Performance? Amount of data required for training? Time required for training?
  • How would that be implemented? Can you write some code on the board.

Even if you don’t know the answer (and why should you? That’s why you’re employing them) good candidates should be able to explain these sorts of things, and it will be obvious if a candidate is struggling to answer.

Simple can be better

Simple technical questions can be surprisingly good at testing candidates. I like to use short maths or programming tasks that are related to an everyday problem. It’s surprising, but these sorts of questions can work really well at establishing a baseline for more in depth questions. (Of course, you probably need someone on the panel who can answer these questions, if you decide to use them)

  • Let’s say you have a list of people’s names and their birthdays. Can you write a programme, in any language, that returns a happy birthday message if it’s somebody’s birthday today?
  • There are a whole lot of simple maths or statistics questions on the internet that you could use. It’s usually worth adding a bit of a twist to these questions, in case they’ve already seen them.

In short, I believe it’s a good idea to try to dig deeper, ask for clarifications, or other solutions.

Data Science: Discovering the Uncharted Value in Your Data

Data these days can become a behemoth of a beast, unwieldy in nature with more tentacles wrapped around different bits of information than you know how to manage. Here are 5 areas where data science can help you leverage uncharted insights from your data kraken.

Data Mettle Data Science Kraken, Be curious about your data and discover its value, five ways to improve your business

1. Building customer loyalty and retention

At the heart of it, your data is the voice of your customers telling you what they want, when they want it, how and why. A data scientist is able to analyse data and from it and explain to the business what your customers are telling you. This information is much more honest than simple customer surveys, as it looks at the patterns in customer behaviour.

Data Mettle Data Science Customer retention and loyalty scrabble board Data Mettle Data Science Customer segmentation street art
Using this knowledge, data scientists can build models to understand customer retention and look at what services, products or practices drive customer loyalty and therefore what the business can put in place to encourage retention and reduce churn risk. Data science can support businesses in identifying valuable customer groups and what types of services or products those groups are particularly interested in.                                                                                                                                                                 


2. Achieving profitable growth with new customers

Data science enables organisations to identify barriers for getting new customers or members and how to remove them. It has the potential to predict what new customers want and identify unmet customer needs. This can give you insight into what your customers may want in the future and how to shape products and services for new markets.

3. Gain a competitive advantage by improving customer experience and optimising marketing campaigns

Data scientists can craft a ‘human journey’ map by interpreting the value in your data. Using your data, they will map how customers interact with your organisation, how you listen to them, what path their journey with you takes.  This knowledge is fundamental in gaining a competitive advantage, because you now have rigorous evidence that underpins how you improve this journey. In this way you are able to give your customers or members a much more personalised experience.

This may be in the form of product recommendations, tailored products and services, or optimised marketing campaigns that target specific groups of your customers with campaigns matching their interests and needs.

4. Automate processes, drive efficiencies and reductions in operational costs

Data scientists start with the business challenge and use this to look at how they can create efficiencies and reduce costs. This could include any aspects of business operations such as streamlining logistics and warehouse operations, putting automated processes in place to manage customer transactions or emails, using image detection to monitor remote sites and much more.

The more challenging and complex the business problem, the more exciting and interesting it is for a data scientist to tackle. Data scientists can add huge value to business operations in sometimes unexpected ways.

5. Reduce fraud and risk using real-time data

Data science can be used to search for anomalies in your data and track potential fraud. Real time data is able to be used so that fraud and risks are detected and addressed quickly.


We’ll talk more about each of these in upcoming blogs with lots of examples.


_data journey

Your data can tell you a lot about your customer's journey. Our services can provide you with the information and tools that you need to match your services to customers.