2 answers
Asked
566 views
I need help on the final step to work as a Data Analyst!?
My name is Jimmy and I have been training to be a Data Analyst for over a year, but I don't feel like I've mastered the career and I don't fail safe to make my own portfolio project, please give me the steps and how to create my own portfolio projects to apply as a entry level Data Analyst, please give me step by step instructions on what my workflow has to be and how to explain it to my job interviewer!
Login to comment
2 answers
Updated
Cory’s Answer
Hi Jimmy, if you've been training for it, you are already a Data Analyst, don't discount yourself just because you're new at it!
Let me offer you a workflow/mindset that can apply to job you have today or in the future, that workflow is called PDCA, which stands for Plan/Do/Check/Act. In this mindset, you never actually master anything, you are instead always improving and perfecting it. It doesn't matter what industry you are working in, managers want you to solve problems for them. PDCA is a workflow you can apply to solve any problem, using data.
Let's play it out in a simple real-world scenario:
You are a data analyst working for a car dealership. The General Manager of the dealership is frustrated as to why none of her supervisors can explain why the average time to repair a vehicle has increased by 30% in the last quarter, upsetting customers. What types of data could be analyzed to answer this question?
-Creating a dataset for all the variables involved in the fulfillment of a vehicle's repair:
-list of all orders sent to suppliers 12 months
-the list of all repairs created for the last 12 months
-list of customer reviews for the last 12 months
After the dataset is built, then it's matter of analyzing the data to look for trends. Questions a good analyst will ask:
-How long does it take for an order for a part to arrive after we send it?
-Is one particular supplier taking longer to send their orders than others?
-How many orders are returned? Are we receiving faulty parts?
-Is there a manufacturer that has a significantly higher number of repairs than the rest?
Once you identify the problems from the data, you then apply PDCA. Let's say after the data analyzed, we learn that Supplier A is not only taking too long to engine parts, they are shipping bad parts.
Plan: Reduce the time to obtain engine parts
Do: Change suppliers Z's fulfillment time
Check: After 3 months, does supplier Z ship parts faster, with fewer failures?
Act: No? Change to Supplier X
The lesson here is to be able to take a problem that has been identified and turning into something you can measure and then fix, and actually substantiate it is fixed, with data. After the measure has identified, you look at all the other data points that can directly impact that measure, and look for correlation. Once you find correlation, go fix that thing in a way that impacts a measure.
Let me offer you a workflow/mindset that can apply to job you have today or in the future, that workflow is called PDCA, which stands for Plan/Do/Check/Act. In this mindset, you never actually master anything, you are instead always improving and perfecting it. It doesn't matter what industry you are working in, managers want you to solve problems for them. PDCA is a workflow you can apply to solve any problem, using data.
Let's play it out in a simple real-world scenario:
You are a data analyst working for a car dealership. The General Manager of the dealership is frustrated as to why none of her supervisors can explain why the average time to repair a vehicle has increased by 30% in the last quarter, upsetting customers. What types of data could be analyzed to answer this question?
-Creating a dataset for all the variables involved in the fulfillment of a vehicle's repair:
-list of all orders sent to suppliers 12 months
-the list of all repairs created for the last 12 months
-list of customer reviews for the last 12 months
After the dataset is built, then it's matter of analyzing the data to look for trends. Questions a good analyst will ask:
-How long does it take for an order for a part to arrive after we send it?
-Is one particular supplier taking longer to send their orders than others?
-How many orders are returned? Are we receiving faulty parts?
-Is there a manufacturer that has a significantly higher number of repairs than the rest?
Once you identify the problems from the data, you then apply PDCA. Let's say after the data analyzed, we learn that Supplier A is not only taking too long to engine parts, they are shipping bad parts.
Plan: Reduce the time to obtain engine parts
Do: Change suppliers Z's fulfillment time
Check: After 3 months, does supplier Z ship parts faster, with fewer failures?
Act: No? Change to Supplier X
The lesson here is to be able to take a problem that has been identified and turning into something you can measure and then fix, and actually substantiate it is fixed, with data. After the measure has identified, you look at all the other data points that can directly impact that measure, and look for correlation. Once you find correlation, go fix that thing in a way that impacts a measure.
Updated
Moiz’s Answer
Data analytics is essentially taking raw data and transforming it so it can be presented to stakeholders.
To get started on a portfolio project, I would do the following:
1. Find a dataset that is interesting to you. This can be retrieved from Kaggle.
2. Check for data quality issues in this dataset.
3. Create a dashboard in Tableau or Looker that shows data in different forms (bar chart, line graphs, etc)
These visualizations should be informative and answer questions that someone could have regarding this dataset!
To get started on a portfolio project, I would do the following:
1. Find a dataset that is interesting to you. This can be retrieved from Kaggle.
2. Check for data quality issues in this dataset.
3. Create a dashboard in Tableau or Looker that shows data in different forms (bar chart, line graphs, etc)
These visualizations should be informative and answer questions that someone could have regarding this dataset!