How did you first become interested in Data Science?
I’ve always loved maths and statistics, and after completing my PhD I knew I wanted a career where I could work on similar technical problems, and be challenged intellectually in the same way. What really attracted me to working in industry over continuing in academia is to be able to work on problems with a tangible impact on the world, and on a variety of applications. I don’t think I had this clarity at the time, but Data Science was the natural choice that ticked all these boxes.
What has been your journey to tails.com?
When I finished my PhD, Data Science wasn’t that established in the UK and I didn’t know it was a career option. So I joined the Civil Service as an Operational Research Analyst, which was a great combination of interesting numerical problems with an important impact on society.
As much as I enjoyed it there, the technical challenge of the work just wasn’t enough for me long term. By that time, Data Science was becoming more established in the UK and I heard about the Faculty Fellowship, giving PhD students intensive training and industry placements on Data Science projects. I got accepted, took a crash course in Python, and then joined the fellowship in 2016. As part of that fellowship I worked at tails.com on a retention project, and I’ve been here ever since!
tails.com’s purpose is to improve the lives of dogs and their owners. How does Data Science play into this and was it a motivating factor in joining tails.com?
Our purpose was definitely a motivating factor, that and there being dogs in the office. After I read through the Trustpilot reviews, and just how many dogs' lives were being improved with tailored nutrition, that sealed the deal for me.
Data Science plays into this by helping make our product and service as effortless as possible. If our customers and their pets are happy, then they will stick with us and enjoy the benefits of tailored nutrition for longer. This can mean anything from working with our blend algorithm and ensuring our selected blends taste good, to working out the right information that a customer needs to see in an email or when they log onto our website.
What are the key characteristics and skills a Data Scientist requires
You need a good understanding and confidence with maths and statistics, the ability to extract and analyse a dataset, using Python or another programming language. But aside from these core hard skills, the key characteristics that will make you a great data scientist are a healthy skepticism, good communication and a growth mindset.
Outside of academia, you are never just handed perfect data with a clearly defined problem, and the needed output is rarely just a model in a simple notebook. You need to be able to work with stakeholders to understand the context, be able to assess if this is something that will really benefit your customers, and agree together what the end product should look like. Having the soft skills to navigate this can be the difference between a useful data product and something that sits unused on the proverbial shelf. The requirements of a data scientist in industry are also constantly changing, so a growth mindset is essential to mean you can adapt to the challenge.
One of the main projects you and your team have been working on at tails.com has been Data Labs - so what exactly is it?
It has indeed and it’s a really exciting project. As a direct-to-consumer business, we’re in a privileged position to be sitting on a gold mine of information on our dogs and their habits. There are already some existing important and valuable studies on dog obesity and longevity that focus on a small number of animals in a very controlled environment. But what we can do, is look at trends in hundreds of thousands of dogs, analyse dog owner perceptions of what a healthy weight for each breed is, and how this relates to their pet’s health and longevity.
The main aim of Data Labs is to use this data to contribute to the understanding of dog health, wellbeing and longevity, and to share these insights with our customers as well as other researchers. This will take the form of some in depth papers and studies, as well as some smaller, stand alone insights. And they’ll all be available from our Data Labs page - https://tails.com/gb/data-labs/.
What are the most exciting projects you’ve worked on here?
Data Labs is definitely up there, it’s great to be able to use our data not just for the benefit of our own customers, but to be able to further knowledge around dog health and longevity. Apart from that, we’ve done some really exciting work with our loyalty team recently on anticipating customers’ needs and then providing them with the most relevant information to help. It’s exciting and rewarding because we can really see the impact we’re having. We’re also using some really cool capability to make this happen, so it’s a great joint effort between multiple teams.
What are the future plans for Data Labs?
We’ll keep the page updated with new insights and papers as and when we do them. We’ve got an extensive data set and there’s a lot that we can dig into. We’re also definitely open to collaboration with others, to improve the impact we can have with our data and knowledge.
A word of advice for aspiring Data Scientists who have just started or wish to start their Data Science journey?
Approach a data problem with curiosity and skepticism, focus first on understanding the data and the problem, and then think about what could be useful and relevant. A lot of value can be gained from applying statistics and data analysis and not everything is a Machine Learning problem.