Hotel Reservation data analysis

From Raw Data to Dashboard

To ensure the data was "Boardroom Ready," I followed a rigorous

- preprocessing workflow:

Data Cleaning: Handled missing values and errors as i get date 29/2/2018 it is wrong date so i remove it

- Feature Engineering (Power Query/M): Created custom logic to calculate Total Revenue and Revenue Loss, ensuring we only counted "Realized" cash for a 100% accurate financial view.

- Dynamic Logic: Developed a "Commitment Gap" analysis by grouping Lead Times to see how far in advance guests are "placeholder" booking vs. actually planning a stay.

Key Insights:

The $11M StoryThe trends were clear:

- The Cancellation Bottleneck: We found a 32.78% cancellation rate. This isn't just a stat—it's a "leaking bucket" where nearly 1 in 3 bookings never reaches the bank.

- The 90-Day Danger Zone: Bookings made more than 3 months in advance are the highest risk. Without "skin in the game," these guests are 40% more likely to churn.

- The Online vs. Offline Paradox: While Online channels drive the most volume, Offline/Corporate segments are significantly more stable and reliable for long-term forecasting