As one of the leading property managers in Bali, Bukit Vista is constantly finding new innovations in transforming and optimizing our 140+ property partners to bring more revenue and guests for their rental properties. One of the ways we try to optimize their revenue is through dynamic pricing, a strategy where we adjust the prices of our listing prices to account for changing demand and other data factors. This was done manually previously by our revenue specialist, but thanks to our brilliant visterns, we’re able to automate this process through machine learning. Interested? Read more below on how they manage to implement this.
Before that, let’s introduce the visterns who manage this project.
A bit of background
A little background on dynamic pricing, it is a program where the Bukit Vista revenue managers looks at the past data of each property using graphs and connect each insight to a single conclusion which is the best pricing decision to optimize the property performance. Imagine writing a mini report manually dozens of times a week! Knowing there’s a huge amount of data that needs to be analyzed each day, the idea of manual work might be too burdensome. As an overview, the mini-reports were made each week for each property. Imagine looking at 5 – 6 graphs for each property, interpreting them, and finally connecting the dots which create a single pricing decision. This usually was done manually, making it very time-consuming and not efficient. Not to mention, increasing the level of human error. Also, with the limitation of how many people can be involved in the project and how far our knowledge management has reached, an innovation of making an automated dynamic pricing occurs. The idea is to reach the same level of efficiency without losing any quality.
To give you an idea, take a look at the figure below. Let’s say we do have a room that we’d like to sell this December 2022. We identify booking opportunities by observing the gaps that exist in between our current bookings on the calendar. A pricing analyst identifies those opportunities as NAB (Next Available Block) that we could sell to generate revenue.
Once the pricing analyst identifies those opportunities, they will apply different pricing strategies for each NAB in order to increase the chance of attracting more bookings. This pricing strategy is called PED (Pricing Event Decision).
After they got paired with each other, they will be monitored and evaluated over days. If the pair successfully attracts booking, they will become PWF (Proven Workflow), meaning it’s a pricing strategy where the strategy itself fits the current market considerations.
Promenade collects NABs from all of the Properties here in Bukit Vista and stores it in a database. From there, all those NABs are learned by a machine learning algorithm called Decision Tree. Overtime, more and more NABs are learned and eventually Promenade comes up with a general rule about how to give pricing strategy based on the data collected.
With that in mind, our pricing analyst just had to give the NABs that they’d like to optimize to Promenade, for example such as shown here below.
Promenade generates recommended pricing strategy (SPED, stands for Suggested PED) for that NAB! This eases up the pricing analyst workload and significantly increases the speed of pricing analysis.
Of course, up until today Promenade is still improving itself to perfection. Overtime, every single data that we supply to Promenade will increase its efficiency and effectiveness to generate a better pricing strategy. This ensures a continuous improvement and a more scalable pricing adjustment for the properties here in Bukit Vista
Inspiring project right? We always appreciate our Vistans and Visterns, who have Inspired Delight and created an impact for our employee, guests, and partners.
We value their talents and see them as a precious asset to our company.
Interested in joining us transform the tourism and hospitality industry?
Contact us at our careers page and see how you can make an impact here also