There are many articles about the skills needed to be a good data scientist or data analyst, but few articles talk about the skills needed to be successful—whether it's exceptional performance appraisal, praise from management, a promotion, or all of the above. Today we bring you a material whose author would like to share her personal experience as a data scientist and data analyst, as well as what she learned in order to achieve success.
I was lucky: I was offered the position of a data scientist when I had no experience in Data Science. How I accomplished this task is another story, and I want to say that I had only a vague idea of what a data scientist does before I accepted this job.
I was hired to work on data pipelines in connection with my previous job as a data engineer, where I developed a data mart for predictive analytics used by a group of data scientists.
My first year as a data scientist involved creating data pipelines to train machine learning models and put them into production. I kept a low profile and did not participate in many meetings with marketing stakeholders who were the end users of the models.
In my second year at the company, the data science manager in charge of marketing left. Since then, I have become the protagonist and have been more involved in model development and project timing discussions.
As I talked with stakeholders, I realized that Data Science is a vague concept that people have heard about, but do not quite understand it, especially when it comes to top management.
I built over a hundred models, but only a third of them were used because I didn't know how to show their value, even though the models were requested in the first place by marketing.
One of my team members spent months developing a model that senior management felt would demonstrate the value of a data scientist team. The idea was to extend this model to the entire organization once it was developed and encourage marketing teams to apply it.
It turned out to be a complete failure because no one understood what a machine learning model was and could not understand the value of its application. As a result, months were wasted on something that no one wanted.
From such situations, I have learned certain lessons, which I will give below.
Lessons I Learned to Become a Successful Data Scientist
1. Set yourself up for success by choosing the right company.
During an interview at a company, ask about the culture of data and how many machine learning models are adopted and used in decision making. Ask for examples. Find out if the data infrastructure is set up to start modeling. If you spend 90% of your time trying to pull raw data and clean it up, you will have little to no time left to build any models to demonstrate your value as a data scientist. Be careful if you're hired as a data scientist for the first time. This can be both good and bad, depending on the data culture. You may encounter a lot of resistance in implementing the model if top management hires a Data Scientist just because the company wants to be known as using Data Science to make better decisionsbut has no idea what it actually means. Also, if you find a data driven company, you will grow with it.
2. Know the data and key performance indicators (KPIs).
At the beginning, I mentioned that as a data engineer I created an analytics data mart for a team of data scientists. Having become a data scientist myself, I was able to find new features that improved the accuracy of the models because I worked intensively with raw data in my previous position.
By presenting the results of one of our campaigns, I was able to show models generating higher conversion rates (as a percentage), after which one of the KPI campaigns was measured. This demonstrated the value of the model for business performance that marketing can be linked to.
3. Ensure model acceptance by demonstrating its value to stakeholders
You will never succeed as a data scientist if stakeholders never use your models to make business decisions. One way to ensure model acceptance is to find the pain point of the business and show how the model can help.
After talking to our sales team, I realized that two representatives were working full-time, manually going through the millions of users in the company's database to identify users with single licenses who were more likely to switch to team licenses. The selection process used a set of criteria, but the selection process was time consuming because reps viewed one user at a time. With the model I developed, reps were able to select users most likely to purchase a team license and improve their conversion rate in less time. This resulted in more efficient use of time by improving the conversion rate for KPIs that the sales team may be related to.
Several years passed, and I repeatedly developed the same models and felt that I did not learn anything new. I decided to look for another position and eventually got the position of data analyst. The difference in responsibilities just couldn't be any bigger than when I was a data scientist, even though I was once again backing marketing.
This was the first time I analyzed A/B experiments and found all ways in which an experiment can go wrong. As a data scientist, I didn't work on A/B testing at all, because that was reserved for a team of experimenters. I have worked on a wide range of marketing-influenced analytics, from increasing premium conversion rates to user engagement and churn prevention. I learned many different ways to view data and spent a lot of time compiling the results, presenting them to stakeholders and senior management. As a data scientist, I mostly worked on one type of model and rarely gave talks. Fast forward a few years and move on to the skills I learned to be a successful analyst.
Skills I Learned to Become a Successful Data Analyst
1. Learn to tell stories with data
Don't look at KPIs in isolation. Connect them, look at the business as a whole. This will allow you to identify areas that affect each other. Senior management looks at the business through a prism, and a person who demonstrates this ability is noticed when it comes time to make a promotion decision.
2. Provide Actionable Ideas
Provide business actionable idea to solve the problem. It's even better if you proactively offer a solution when it hasn't already been said that you're dealing with a primary problem.
For example, if you told marketing: “I noticed that the number of visitors to the site has been decreasing every month lately”. This is a trend that they may have noticed on the dashboard and you didn't offer any valuable solution as an analyst because you only claimed an observation.
Instead, examine the data to find the cause and suggest a solution. The best example for marketing would be: “I've noticed that we've had a decrease in the number of visitors to our website lately. I found organic search to be the source of the problem due to a recent change that caused our Google search rankings to drop.". This approach shows that you have been monitoring the company's KPIs, noticed the change, investigated the cause, and provided a solution.
3. Become a trusted advisor
You need to be the first person your stakeholders turn to for advice or questions about the line of business you support. There is no shortcut because it takes time to demonstrate these abilities. The key to this is consistently delivering high quality analysis with as few errors as possible. Any miscalculation will cost you confidence points, because the next time you provide an analysis, people might wonder: If you were wrong last time, can you be wrong this time too?. Always double check your work. It also doesn't hurt to ask your manager or colleague to look at your numbers before submitting them if you have any doubts about your analysis.
4. Learn to Communicate Complex Results Clearly
Again, there is no shortcut to learning how to communicate effectively. It takes practice, and over time you will get better at it. The main thing is to identify the main points of what you want to do and recommend any actions that, as a result of your analysis, stakeholders can take to improve the business. The higher up the career ladder you are in an organization, the more important the ability to communicate. Communicating complex results is an important skill to demonstrate. I have spent years learning the secrets to success as a data scientist and data analyst. People define success in different ways. To be described as an "amazing" and "star" analyst is a success in my eyes. Now that you know these secrets, I hope your path will lead you faster to success, however you define it.
And to make your path to success even faster, keep the promo code HORNBEAM, for which you can get an additional 10% to the discount indicated on the banner.
More courses
Recommended Articles
Source: habr.com