How to get real-world experience while providing actual value
For aspiring data scientists and data analysts, it can be frustrating breaking into the field without ‘real-world’ experience. In prior articles (here and here), I talked about creating your opportunities. You may want to look around in your current job and see how you can apply your data science interests and skills. I want to cover some projects such as process automation and text analytics that may help you gain that experience in your current role. In this article, I will review process automation and why you should do it.
Business Process Automation (BPA)
Business Process Automation (BPA) is any process that streamlines workplace processes and possibly reduces costs. Robotic Process Automation is a type of BPA and uses machine and AI techniques. We will focus on BPA to start.
How automation helps your team
If your team finds itself bogged down in repetitive manual tasks, there are opportunities for BPA. If you can reduce the amount of non-value-add work, the staff will be freed up to work on other valuable projects. Some examples of such scenarios include:
- Each week, the project manager has to clean up the formatting in three spreadsheets, calculate KPIs and merge them together into one report. The dates are usually a mess, and why does accounting always CAPITALIZE EVERY WORD?
- Each month, a manager is creating a spreadsheet with departmental KPIs and loading it to Tableau for her boss’ staff meeting.
- Each Wednesday, a data analyst is manually pulling data from different reports and loading them to another database, such as Access or AWS.
- Each morning one employee arrives early to put together the number of customer calls from the day before. The manager likes to have it first thing in the morning, so if coverage is needed, if the employee takes time off.
Do any of those sound familiar? No doubt. The opportunities are all around you.
How implementing automation helps you
What does this have to do with Data Science? Being able to pipe in and format data in large volumes is part of the job. The less manual work you spend on repetitive tasks, the more time you have to do analysis. While this skill alone won’t get you to the top of any interview lists, it is a solid skill to have and will your work life better. It is generally easy to implement and has a good Return on Investment (ROI).
What you can do
Let’s review the first example task. For all the glitz of AI, most departments still have excel spreadsheets floating around with essential data. You can write scripts that will systematically clean up department spreadsheets and merge them. You can schedule it to run weekly before the staff meeting and have it emailed to the manager to review. One step better may be to set up a Tableau report that gets updated automatically. What manager is going to turn down an offer to have their most annoying tasks automated?
A novice-level python script can take care of this entire process up until Tableau. You can knock it out in an afternoon, making your effort risk a few hours.
There are many great tutorials online on how to code this. My recommendation is to keep the code simple and add comments.
Make sure you test your script and have it validated by your manager! I wish I didn’t have to say that, but best to put it out there.
Provide evidence of value ($$$$)
Once you have completed the automation, calculate the ROI. Each company does it a bit differently based on the value per full-time employee (FTE), so you might have to inquire with your manager or finance. You can start translating your skills into revenue and savings.
There are numerous reasons to calculate the savings as a result of your work. If you can prove you are freeing up your team to work on higher priority tasks, requests might come to work on more complex tasks, stretching your skills. You can also use those calculated savings in your performance review and on your resumes.
Here is a sample ROI that I created. The definition is sourced from this blog by Big Sky Associates.
Opportunities to automate business processes are all around you. Look for a low-risk process to start with. Keep it simple. Test it. Calculate the ROI. Put it all together and build your confidence and your personal brand at your company. You don’t have to wait for the official title to do data science.