With data scientists emerging as one of the most sought after positions, organisations across the world are looking to increase their data science capabilities. This has resulted in a deluge of jobs, with demands coming in from new startups as well. But do they need to hire a full-time data scientist early in its life cycle?
Indeed, there may be great uses for hiring data scientists, especially given that they can gather insights that can significantly contribute to overall business success. But without adequate customers and a proper data infrastructure in place, data scientists can struggle to find value at early-stage startups. In fact, the same holds true for well-established firms with a small customer base.
However, it is important to note that small companies should nevertheless foster a data-driven culture, and seek to understand how that approach could be integrated with their strategy. This can eventually funnel into a real requirement for a data scientist, but at a nascent stage, what they would require instead is a ‘data person’.
What Does A ‘Data Person’ Do?
Give them any title you like, but early-stage startups may do well to appoint someone who can make sense of the little data they have, help build proper data systems, and run experiments on cultivating a data-driven approach. Until they have accumulated enough customers and data to justify the cost of hiring a data scientist, they need to make do with the skill sets available.
At this stage, it is acceptable to not have a full understanding of what the startup needs or how they can leverage their data to make decisions. Instead, the focus needs to be on how to set up the company to be data-driven tomorrow.
In case you still need convincing, answering these three questions will give you a clearer picture of whether or not your startup should freeze its data scientist hiring initiatives:
Do You Have Enough Data?
Many of the techniques that data scientists’ employ often require reasonably large samples of data points. This is especially true for deep learning, which relies on massive amounts of training data.
But even if you have enough data, it first needs to be collected and cleaned — a process that is very time-consuming. This is because data is typically accumulated and stored in multiple formats and at different places, making it challenging to convert it into a format that is usable. And this is not all. Only once the data is prepared, can it be modelled according to the needs of the company in order to extract insights.
This means that although you may not need a full-time (and expensive) data science resource at this point, there is still valuable work to be done to create an environment, where data science can thrive in the future.
What Will This Data Scientist Do Once Hired?
If you have no clear answers to this question, you might want to reconsider your hiring strategy. It is true that the variability in skills in data science is substantial, but you need to have more than just an idea of what value they can bring to your startup.
A better approach would be to upskill existing employees/analysts such that they can perform basic data jobs, like forecasting the KPIs they report on. This presents them with the opportunity to learn on data they are already familiar with.
Will Your Data Scientist Have Adequate Support Networks?
If your data scientist does not have the support of senior resources in the team for guidance, it is likely that they will not be able to perform to their full potential. While formal data science programs would have helped them with the technical skills, it is not enough to build a nuanced understanding of businesses.
What is more, inexperienced data scientists or data science graduates may be used to solving pre-established problems on clean data, which is almost never the case in the real world. This can be tackled with a little help from seniors.
Therefore, if you must hire a data scientist, build a team of data science professionals with varying levels of experience. This can also help you avoid costly mistakes that come with incorrect analysis. However, this problem cannot be completely mitigated by building a whole team of data science experts. They will still need the backing of people at the executive level to get their insights translated into helpful applications.
No matter where they stand, all organisations need to carefully consider what kind of data scientists their company needs — if at all. However, early-stage startups should remain lean.
The landscape for hiring good talent in data science has become increasingly competitive, but this can be offset with smart decisions on when, who, and how to hire. The success of your data program will depend on careful consideration of your requirements before making any hiring decisions.