Let’s say you’re a marketer who’s got a passion for numbers, a bit of a stats geek, you love an illustrated workflow, and you’re having a change of heart on your chosen career – you now want to retrain as a data scientist. The fact of the matter is, there is a growing global demand for competent data analysts and PhDs within data science disciplines, but not enough candidates to fill the roles.
So, how do you transition into a career as a data scientist from more of a marketing-based role, and what temperament should you have?
Marketers seeking to transition their career could be of great strategic value to a company. They can offer insights into how to gain market advantage, given their perspective on wider business processes and often have a very holistic understanding of each of your customer segments.This is something that’s often lacking for most data scientists, many of whom arrive directly from academia, with little real-world ‘on the job’ experience.
Fundamentally a new type of job shift is emerging, blending the skillsets of a marketing expert to that of a data scientist.
The “engagement scientist” can be seen as an evolution of the data scientist’s role. We can — and should — see engagement science pick up steam as a discipline in the next couple of years, in my view. And what special skills or attributes does the engagement scientist need?
Evolving role – the “pure” data scientist becomes the engagement scientist
Growth hackers may practise engagement science, but by no means should growth hacking be equated to engagement science. Yes, growth hackers and engagement scientists both bridge marketing, product and engineering disciplines, but there is a fundamental difference. Engagement science connects its measurements (such as level of engagement with an email through open rates) directly with tangible marketing outputs through statistical prediction, as opposed to just being based on commonly-held beliefs or a marketer’s intuition.
An example: a growth hacker might measure the last six months of engagement with a prospective customer list based on email opens or call-to-action clicks, and then choose to test with a series of specialised emails to a smaller more engaged segment to bear out a willingness to move down the funnel. In other words, they are reacting to what has been observed in the past. An engagement scientist, on the other hand, would build a prediction model for what success means, then send each email to the specific target that maximises its predicted value.
Engagement scientists use a statistical model, not just heightened intuition or a “best practice”, in order to see through those hypotheses which can be borne out by the data.
Are there disadvantages moving from a marketing role to the data science team?
In terms of expectations for data scientists entering most companies, if you are currently a marketing analyst or in a business development role, you may be at a slight disadvantage in comparison to other prospective data science candidates.
Certain companies recruiting for a data scientist will expect a doctorate-level certification in Computer Science, Statistics or the physical sciences at the very minimum. That’s not to say that you cannot make the leap with transferable working skills, but it’s not as commonplace. That said, if you have a fundamental grasp of product marketing, you are on your way to bridging that gap.
The most common disadvantage for senior marketers turned data scientists is likely their dependence on marketing dashboards or marketing automation tools as opposed to being able to directly code whatever type of analysis they need. Making this transition would demand a very creative and flexible marketing analyst. Data Science is definitely not just about learning to use “data science software”, there is a more detailed understanding required.
So.. how to make that leap?
A blend of statistician, computer scientist and creative thinker, engagement scientists – just like the traditional data scientist – must have an innate ability to:
- think in software and/or product development terms
- have the skills to collect, process and extract value from diverse sets of data
- understand, visualise and communicate findings to non-data scientists
Mostly, it comes down to a particular company’s needs and what kind of insights need to be drawn from the data.
If you work at a much larger organisation, think of the engagement scientist’s key role as acting as a bridge between data science, engineering and marketing departments so that no silos occur. By contrast, if you work in a more nimble start-up environment, the engagement scientist’s role must be flexible: you might have to wear several hats, start out in customer experience or growth hacking, but show an aptitude and willingness to grow and have a knack for tackling large data sets on your company’s behalf.
Both have their distinct advantages or drawbacks, but, if you’ve got the basic drive, understanding and willingness to adapt and learn in a new field, with the right supportive environment, it may just be the career leap worth taking.