Data Visualization

Blog of the Data Visualization & Communication Course at OSB-AUB

This is my favorite part about analytics: Taking boring flat data and bringing it to life through visualization” John Tukey

When Do Coffee Lovers Really Show Up?

When Do Coffee Lovers Really Show Up?

If you’re Lebanese, you know that coffee is more than a drink. It’s a companion for all occasions; from kick-starting your morning with an energy booster, to a beverage to enjoy while having conversations with family and friends, and especially during late-night studying. It’s how we wake up, how we host, and how we quietly share connection and love.

With that in mind, I wanted to look at coffee from a different angle: through data. Using a dataset of transactions from a café, I explored:

  • Which drinks are most popular

  • How consumption changes across the time of day

  • Which hours are the busiest

  • And how daily revenue fluctuates across the week

The goal is simple: if coffee is so central to our culture, when exactly are we showing up, and what are we ordering?

Why does this matter? Café owners find it hard to know exactly when to staff up and which drinks to push.

This visualization ranks drinks by total revenue. A few patterns stand out:

  • Lattes and Americanos bring in most of the money, with a balanced consumption of each drink throughout the day.

  • Other drinks, such as Cortado and Hot Chocolate are also consumed but at much smaller level compared to their counterparts.

Despite having many options to choose from, individuals tend to enjoy the classics. These are the drinks that keep the line moving and the espresso machines busy.

 

Next, I looked at revenue by hour of day. This gives us a clear view of when people frequent cafés the most. The pattern is very typical:

  • A strong morning rush between roughly 8:00 and 10:00 AM as people grab coffee on the way to work or university.

  • A smaller afternoon bump, around 3:00 to 5:00 PM, when people need a second boost or meet friends.

  • Sales gradually drop off into the evening.

These are the hours when extra baristas would be needed the most, and popular drinks should be front and center to keep service fast.

Time of day isn’t the only dimension that matters. I also wanted to know:

Are some days of the week more predictable than others?

This boxplot shows the distribution of daily revenue for each weekday:

  • The middle line in each box is the typical (median) daily revenue.

  • The height of the box and whiskers shows how much that day’s revenue can fluctuate.

From this view:

  • Weekdays tend to have higher median revenue than weekends.

  • Some days, especially mid-week and Friday, show taller boxes and more high outliers, meaning they can be very good or just average.

  • Weekends (particularly Sunday) are a bit calmer: slightly lower but more stable daily sales.

So if you’re planning staffing or inventory, weekdays are where you need more flexibility, while weekends can be managed more steadily.

Because most revenue comes from weekday morning and late-afternoon peaks driven by a few core drinks, we should staff and promote around those windows to serve faster and sell more