Celonis Challenge: Mamma Mia Pizzeria TAMU Datathon 2021
What it does
This project uses Celonis analyses and figures to identify drawbacks in the pizzeria's processes. Then, a data driven approach is used to propose improvements to increase customer satisfaction as well as profitability.
How we built it
Challenges we ran into
The Celonis grammar of graphics is still growing, so I had to make tradeoffs in terms of the number of charts and the expressiveness of charts.
Accomplishments that we're proud of
The Celonis webapp as well as the submissions were fun to make.
Local pizzeria owner Giovanni is confused why his generation's old “original recipe” pizza receives 0 out of 5 reviews with comments like “never ordering from here again”. The small-business owner will be glad to hear that a Celonis analysis can pinpoint minor and major process issues that prevent his pizza from getting the 5 star reviews they deserve. Many dissatisfied customers complain about the seemingly “several hour” wait times for their pizza. Customers calling in for their order are surprisingly likely to face a large delay. 171 out of 1117 of those orders follow a variant where 5 minutes after a pizza is baked, it needs to be prepared and baked again, resulting in an **18 minute delay** on average. Another 154 callers were called by the pizzeria because of confusion between when the order was placed and when the pizza was prepared. This mix up caused an **8 minute delay** on average and can be addressed with an __improved system of communication between the counter and kitchen__. Giovanni should work with his chefs so that the improved system fits every employee’s needs and prevents a delay and wasted pizza for 29% of phone-in orders. Only 19% of customers use the website, likely because it is a time consuming process. The process of starting and receiving a website order routinely takes **over 10 minutes**, a process that could be __significantly streamlined by investments into a new website__. Any efforts to reduce wait time should simultaneously improve customer satisfaction. In fact, patrons with a 3+ star experience waited an average of 32.0 minutes, compared to 49.9 minutes for dissatisfied customers. Using Celonis analyses and Bayes’ rule, we can predict that 61% of customers eating pizzas made using a conformant process will be satisfied. Conversely, **71% of customers whose pizza was delayed due to nonconformance would leave a poor rating**. Conformance analyses show how much of a difference a __streamlined, conformant process devoid of unnecessary delays would make for improving customer satisfaction__. The demographics of patrons reveals that all groups are approximately split on their rating of the pizzeria. Students make up approximately 57.7% of customers yet are **satisfied only 48.7% of the time**. To boost the overall rate, Giovanni should __advertise his new improvements around college campuses__ so that __more core customers have a positive impression of his pizzeria__. The insights revealed by Celonis analyses should significantly reduce throughput time and increase customer satisfaction. With this boost, Giovanni should expect an influx of diners ordering from his restaurant. This wave is a perfect time to increase profit as well. The analysis reveals that **profit and fulfillment time are uncorrelated**, which means that improvements to order time will not raise profit on their own. Instead, analyzing cost factors reveals that online and phone orders are most profitable. __Delivering pizzas using an improved, conformant process should be a priority__; orders with delivery scooters, phone bill, and delivery guy 1 average profits of $8.54, $8.49, and $7.74, respectively. On the other hand, the **average profit of orders with waiters as the primary cost factor is $6.03**, suggesting that **dine in meals are less profitable than website or phone orders**.