Turning algorithms into visualizations to bring data to life
The Sensibill Barcode Report shows how we can make actionable impacts with data
Here at Sensibill, we are on a mission to humanize banking. The Sensibill Barcode Report is the perfect example of how to use data in a meaningful and helpful way. When creating these monthly reports, the process can be broken down into three main parts: exploration, collection and interpretation, and visualization.
Exploration 🔎
We start by going through the data that we have. We pick a theme or topic (which is usually the monthly Barcode Report topic) to narrow down the vast amount of data we have, although there are still tons of numbers to go through! We look at the SKU level data to form grouped topics and main trends that define a purchasing behavior. Our goal here is to connect purchase activities to categories like demographics, time, location, themes, etc.
Collection & Interpretation 📈
Next, we look at the trends over time and cluster data, which we group by behavior trends or targeted demographics. In this phase, I determine which algorithms can be applicable to the data to help us reveal trends. To keep it simple, the algorithms are a standard set of math equations that can have meaning and interpretation if applied properly. I am using traditional algorithms that wouldn’t be used in a lot of cases, so it’s a lot of trial and error, and I get geeky excitement when the formula finally works! 🤓 👏
The algorithms can be intimidating, so it’s my job to make sure we can give them a friendly face. 😊 This is why the interpretation is so important. We are taking crazy amounts of numbers and formulas, and turning them into words and values that have meaning. This leads perfectly into the final step...
Visualization 👀
This is how to make the data digestible and is one of my favorite parts of the process. It is arguably the most important part of this process and is where I feel as though I am bringing data to life! To put it in perspective, showing a number is like 2D, but putting it in a visual format gives it that 3D, all-encompassing look. Here are some examples:
Image from the October Barcode Report about Thanksgiving trends in the U.S. and Canada.
Image from the July Barcode Report about pet spend in the U.S.
Image from the May barcode Report about Home improvement spend in the U.S.
Globally, we still have a long way to go when it comes to data. We probably aren’t even using 1% of the data available in this world to its full potential. If we use the same three categories as above, we are still in the exploratory stage of data analysis and collection.
I hope in the next 10-15 years, we can use data to its fullest so we can predict things before they happen. Like the 2008 financial crisis. If we could predict even a year earlier, we could have changed some of the negative outcomes.
Using data for good; that’s my hope. 🤗 And that’s what we’re trying to achieve here at Sensibill.
Our next Sensibill Barcode Report comes out in December! Want to get it directly in your inbox? 📧 You can subscribe here!
Priyanka Mukherjee
Product Data Analyst
While completing her Master of Business Administration, Priyanka fell in love with data. At Sensibill, she uncovers customer spend insights through algorithms, data analysis, and visualization. Outside of work, you can find Priyanka trialing a new recipe, watching a thriller on Netflix, or teaching her cat to do some new tricks!