top of page

How Data Science Helped Walmart Predict Sales During a Hurricane

The weather. We can't control it, and it affects our lives every day. Usually it behaves, but often Mother Nature reminds us who has the power. In those circumstances, our world is shaken and our norms broken.

During these irregular times, how do you ensure to operate at the highest level, keeping your business running while helping people who depend on you.

Hurricane Francis was bellowing in the Caribbean, on a direct path to Florida. Walmart started to wonder how they could better improve their supply chain before large storms. What items should they increase in inventory? What would people want?

Armed with loads of sales data from previous years, Walmart looked at previous hurricanes and their effect on store sales. What was unique in their sales during these tragic events?

They wanted to start predicting what's going to happen, instead of waiting for it to happen.

Of course, of course, there were the obvious answers high sales of umbrellas, flashlights, and bottled waters, but that's not all Walmart found.

  • Strawberry Pop-Tarts sold at a rate 7x their usual before hurricanes.

  • The most sold item before a hurricane? Beer!

Walmart sent beer by the barrel and thousands of boxes of those breakfast delights in preparation for the storm, not only giving the customers what they wanted, but making a buck by doing so.

For technical details, see this Kaggle competition:

604 views0 comments

Recent Posts

See All


bottom of page