Machine learning has some powerful capabilities when applied correctly to a business objective. We started a journey last year to build a dynamic pricing tool to transform how the Motorcoach industry operates. In this blog, we’re going to discuss some of the benefits we discovered while building a dynamic pricing tool.

Initial Challenges

There were several major challenges with building the dynamic pricing tool but we will focus on the data constraints for this blog. Majority of the challenges existed in the consistency and how the data was stored. There were many instances in which we discovered the same data but with different labels. Our main goal was extracting the most relevant data to use for our machine learning algorithms.

Automating Decision Making

One of the biggest challenges the Motorcoach industry currently face is how the price is created. Currently, a customer visits the site and books the days they need a bus rental. The sales team then looks at past prices and develops a price to send back to the customer. As a result of layers of manual steps, this process can take up to 24hrs to complete.

We wanted to build a tool that would solve this 24hr lag and return prices in seconds instead of hours. Once we identified the business problem, we started gathering data. After the data preparation, we built a machine learning algorithm that allowed us to identify the peak and slow periods for booking a bus. In conclusion, automating the decision making improves customer engagement by reducing the time to get results.

Testing The Tool

During the testing phase, we discovered several other ways to build the tool that proved to be effective as well. Over the course of several months, we created multiple versions of the dynamic pricing tool. We noticed throughout our experiment, certain times of the year and week had higher booking activity. The owner was well aware of the seasonality in the business which helped guide us in the building process.  Our end results was a tool that noMachine Learningt only predicted future trends but would automatically increase or decrease the price for a given day.

End Result

Our final solution resulted in three versions of the dynamic pricing tool. We built a series of tools that can solve a variety of pricing challenges most companies face. The tools we created were demand-based, inventory, and time-of-day pricing. These tools can be tailored to solve a variety of pricing challenges.

Alex Brooks is the founder and CEO of AE Brooks, LLC (d/b/a,Entreprov), a Seattle-based firm that builds custom predictive analytics and automation tools to enhance a company’s performance and decision making.