How to be your own data scientist

Crisp Product Manager Michelle Hood demystifies data science to help CPG teams answer key questions and make more informed business decisions.

“Why should I care about data science?”

There are certain people who have a talent and passion for statistics and data science. As for the rest of us, we quickly put that college statistics course aside and jumped into our careers. But demand for data science is exploding in our data-driven world, and analytics is now a skill set that each of us needs to hone. Think of it as the new Excel for the 21st century.

Post pandemic, Data Science has become a buzzword along with Machine Learning (ML) and Artificial Intelligence (AI). Business leaders are accelerating data-driven strategies within their companies to adapt to shifting consumer demands and omnichannel shopping opportunities. Companies (and their employees) are left with the decision to either join or be disrupted by this revolution.

Let’s start by understanding what data science actually means: the ability to “design and construct new processes for data modeling and production using prototypes, algorithms, predictive models, and custom analysis” (Northeastern University). Data science draws on skills in mathematics, computer science, and domain knowledge in addition to traditional statistics.

While being a data scientist is not a calling for all of us, the ability to harness data can serve many of us in our everyday working lives, whether that’s marketing, sales, or operations. We can, and should, embrace new data processes to get actionable insights and make more informed decisions (without getting another degree). In this post, I’ll share how you can embrace data science by following three not-so-scary steps.

1. Start by asking questions

You don’t need to be a trained data scientist to harness the power of data in your daily actions. However, your approach to data does need to shift from simply accepting the same old data set to asking deeper questions and finding the necessary data to answer those questions. This is the first step towards the paradigm shift of becoming an empowered data champion. 

Start by assessing the data you currently use. What does this data help with, and what are the gaps that you often fill with assumptions? Ask yourself what data would help fill those gaps. Are there any internal or external sources for that data that would be helpful to incorporate? Lastly, go ahead and brainstorm by asking what data would truly make your job easier and yield better results for your organization. What seems impossible to obtain from your existing data?

For example, when I oversaw 100 SKUs for a large CPG company, I started each day by printing several pages of spreadsheet data to review sales versus inventory for each SKU (a big waste of paper). I then had to comb through emails to pull information about product complaints from our Customer Support team, and ended the day meeting with our demand planner for inputs on production velocity accuracies. I was spending half my day just seeking data and was frustrated that important information from marketing and sales were only accessible via meetings. Feeling restricted from doing my job well, I began to ask questions on how to improve our data systems and get better insights.

2. Fill in the gaps

Now that you’ve identified the gaps in your data, your next mission is to seek out the data that you need. Start searching through your organization internally, and from there, see what data you’ll need to obtain externally. It’s a worthy investment to find and get to know the IT contact connected to your internal data systems. This contact should be able to provide an overview of the constraints and opportunities that lie within them.

In my journey to improve my product portfolio data, I followed the same steps: I identified key data points that were missing, worked with my cross-functional teams to determine what else we could obtain, and then developed a plan to capture and feed the data into a new analytics tool. I then built visualizations fed by new automated data flows. These efforts resulted in a better understanding of our distribution and production velocities, where our largest sales were coming from, emerging product trends, and issues down to the production day/location/shift. It also offered line-of-sight into aging products down to the case in each warehouse, enabling us to take action to prioritize that inventory.

3. Time to analyze

You’ve gained access to some new and exciting data…now what? Before you panic, go back to the assessment you did in the first step. What open questions do you have that this data needs to solve? After revisiting your objective, explore how you can organize your new data to answer those questions. This is where your time investment with your IT contact will be handy, as they should know what tools you have available to do this. Whether you use systems that enable specific data queries or have access to data visualization tools like Power BI or Tableau, find the right fit that reveals meaning in your new data.

Another option is to partner with Crisp. Crisp aggregates retail and distributor data to provide  in-depth, actionable dashboard visualizations that make insights easy to uncover. Crisp’s own data scientists have implemented machine learning and predictive analytics to help you resolve unanswered questions, from detecting possible voids to identifying new locations for distribution. This frees up time for you to focus on what you do best and take action based on your insights. Additionally, Crisp offers your organization multiple opportunities to export data into the tools you already use, like Excel, PowerBI, Tableau, and Snowflake.

To learn how Crisp can help you be a data champion, contact us today for a demo.

To keep up with other industry insights and data best practices, subscribe to the blog.