When browsing the internet , you can find many articles about the skills needed to be a data scientist or a data analyst - and lot's of them have valid points , but there are few that tell you the skills needed to be a successful one...
Things like how to get a raise, how to get an exceptional performance review or praise from management, a promotion, or all of the above - those are things that I have found less information available when I did my first steps in the data space.
Today I’d like to share my firsthand experience as a data scientist vs. a data analyst and what I learned to become successful.
I have started my career as a data engineer and shifted to data science. I was fortunate enough to be offered a data scientist position (PM oriented) without any experience in data science.
How I managed this ? Let's just say that I knew how to build datasets for ML models and data pipelines as a Data engineer :) , but my point here is that I only had a vague idea of what a data scientist did before accepting the job. Afterward I lead analytics and data science teams - hence the data analysis experience. Your first year as a data scientist would probably involve building data pipelines to train machine learning models and deploy them in production, sometimes working with team members on a research as an assistance and mostly data cleansing.
So... let's talk about how to become a successful data scientist !
First - take ownership
The first period shouldn't be handled in the back of the scene! What I mean by that? Usually in your first position you stick to tasks like I have stated without any strong will to be involved in meeting with the stakeholders (product , marketing etc.) and engage with (sometimes) the end users of your deliveries. This is a great way to get To that , either you step up and take the opportunity or someone quit the job and you get the ownership, which one you prefer? This will improve your project management skills , timeline management , your ability to better understand the business problem and needs when creating a data based solution and more important - create awareness for your data science projects around the company. It defiantly worked for me - as I interacted more with stakeholders I realized that data science was a vague concept that people heard about but didn’t quite understand, especially senior management.
Before that - I have built several models that most of them left un-used due to the fact I didn't know to show the value.
Next - choose the right company
The interview part shouldn't be one sided (from asking questions aspect) , this is your chance to get valuable information about the company data culture , if they are using ML at all and if so , what is the scale.
Also - you would like to know if they have a proper ML infrastructure to plug and play , otherwise your day-to-day will be more off cleaning data and less developing solutions and bringing value.
One tip that I have learned - if it's the first data science hiring - handle with care!
If the company is data driven by nature than you will grow as the company grows , but the hiring is only made because the company wants to show they are using data science and have no clue about it - stay away.
Metrics KPIs and Trends - Know them by heart
So as I mentioned before , I have started as a data engineer and build lot's of analytics datamarts that created using the raw data and included all kinds of metrics and calculated features.
That helped me when I was a data scientist as I was able to find new features that improved the accuracy of the models.
When I overviewed the model campaign results I was able to present that the model generated high conversion value and high click through rates which are important metrics , but also the way stakeholders understand and evaluate value.
Adoption and Value are co-exists
Your success as a data scientist is pending by the stakeholders making decisions with your solutions , adoption is critical. To ensure the adoption of your model you can start with finding the needs and pains of the business and translate it into a solution that can help.
It happened to me with the growth and sales teams , they were looking into the CRM database to find free customers that might be a good fit into enterprise subscription.
They were browsing a large database and afterwards comparing the finding to a BI report that aggregate the customers activity in order to find signals.
I have used several of their criteria's and added some more features of my own in order to create a model that scores the potential of customers to upgrade.
This translated to a more efficient use of time by increasing conversion rates for a KPI the sales team could relate to.
Afterwards... I have moved on with my career and got more into data analysis.
There were few major differences comparing to data science position - and I tell about it soon.
As a data analyst you are working with stakeholders on a daily basis , your analysis get used on a daily basis to take decisions upon , your contribution is critical for tactical action and you are doing things that usually you are not doing as a data scientist - such as A/B tests.
As a data analyst you are dealing with things like improving conversion rate , user engagement and churn prevention , and in the first time I have learn to better look at the data , compiling results and presenting them to stakeholders and senior leadership.
As a data scientist I had mainly worked on one type of model and rarely gave presentations.
So let's talk about the things needed to be a successful data analyst
Master the data storytelling skill
As a stakeholder - I want to hear about the overall impact that a campaign or any other activity might create.
Why is that important?
As a data analyst you need to present an holistic view of the metrics that you are overviewing
(hint : don't look at metrics in isolation) that will allow you to identify areas that influence each other.
Senior leadership looks at the business from a bird eyes view and a person demonstrating this ability gets noticed when it comes time to decide on promotions.
Be the favorite expert
You need to become the first person your stakeholder goes to for recommendations or questions about the business line you support. It takes time to demonstrate these abilities - so take the time as there are no shortcuts. What would you need for it?
Keep delivering high quality results and analysis with little to none mistakes - as any error will decrease credibility for the next time analysis.
QA and double check your deliveries - you can use the "Jackie" test to get feedback from others who are not your stakeholders
** The "Jackie" test refers to a person who has no prior knowledge of what you're about to present and provide a genuine feedback about your key points clarity
Make your insights actionable
Pointing out problems it's great but as the person who holds all the data points it is expected to have solutions to those insights you have brought up , better do it by yourself rather than get told to do so.
Let's say that you have analyzed the product engagement and noticed a decline in the monthly goal/funnel completion - saying that alone might not be enough as this information might be available to all , and regardless - there is no value here as you stated the obvious.
Instead you can track back the trendline and look for anomalies or changes in the patterns , and connect the dots to find the reasons for that trend.
For example - I noticed a drop on our engagement funnel and increasing % are not completing the goal , and that is due to large group of safari browsers users... This demonstrated you were tracking company KPIs, noticed a change, researched the cause, and offered a solution to the problem.
Reduce the clutter off your results communication
Learn how to keep the complexity at your end and communicate clear results to the stakeholders.
It's a matter of practice and some more practice.
The key solving that is to focus on the pains/business problem and the main points you want to talk about , and afterward - your actions and recommendations.
The more senior you get in the organization the more important it is to communicate well. The ability to convey complex results is an important skill to demonstrate.
It took me a while to learn all those important lessons and secrets.
Best thing is to define for yourself what is success and draft a plan to achieve that.
I hope those tips will take you there sooner!
Commenti