Revenue Management is More than Numbers
Revenue management is finally catching on in the hostel space, and hostels that don't use it are practically leaving money on the table. We all know by now that you have to change your prices frequently and share data to really get conclusive results. You have to price strategically to grow your demand, keeping a steady pace of bookings coming in through multiple means of distribution. But that is all looking outward, and it (the market and the channels) is old news. It is time to move your strategy into something new. You should now be looking inwards and seeing what insights your internal data can tell you.
To best illustrate this, we'll talk about how successful hotels organize and execute their data, I mean revenue management, to boost the most tRev. The focus of proper revenue management is on the total guest spend or tRev, not just the price for their bed night. This includes their food and beverage (F&B) and other ancillary revenue such as travel sales and services. For hotels, it's the bar, restaurant, housekeeping, entertainment (spa, golf, media, etc.), and yes, you will need a system that tracks that data per guest account to get the most out of RM. The best hotels to look at in this regard would be resorts—think beach or ski—where there are tons of options for the guests to indulge.
Each purchase a guest makes is a data record that becomes part of their folio. In that record, you have many data fields that at least record the when and where. You can imagine each purchase as an item in a shopping basket. Looking at the total purchases of the guest is called affinity or basket analysis (and for the techie). Basket analysis is a popular data mining and business intelligence term. The focus here is to get more guests to add more products into their baskets, which will mean more revenue for you.
By combining those transaction records for each guest, you can create a statistically-based tree graph showing what decisions happen before or after other decisions, commonly known as a decision tree. You can see what events and purchases lead to other purchases, etc. Now we're doing data modeling. If you add or remove a decision, how would it affect the purchases down the line? The larger resorts even go further, combining the trees of many properties to create a random forest, but that is quite advanced.
Pretend you are a resort owner. Since it is winter, imagine you are a ski resort owner. Say you sell day passes to the lift, warm beverages, meals, and a sauna. You can make a tree and see if guests that have a hot chocolate after 6 PM usually have a couple of drinks and then go to bed because they are exhausted. You can call that a bottleneck in your tree. Now imagine you offer a special on hot beverages at 3 PM, and guests return early and not so exhausted. They have a coffee, then go into the sauna, and afterward have a meal and some drinks. That can lead to a major boost in revenue per guest.
This is what major resorts and hotels are doing—looking deep into their operations and how they can be more welcoming to you and your money. What does this mean? It means you use data to back up your guests' behavior. You are understanding your guests statistically. Sure, many hostels don't even obtain enough data sets to begin modeling, but you can still think of and test small changes that could influence your guests' behavior and average spend.
Even if you are a small hostel, you need to focus on what your guests are purchasing, when, why, and how you can deliver that service to them. Larger hostels and hostel groups can pool their resources and conduct tests as well. Now if you have the bed count—or better yet, the headcount, because private rooms matter too—and the right tech in place to record the data to play with, you might want to investigate this further. If you know, please make a comment. Many major hotels and resorts hire data scientists to do it directly. You can find a data science student (most likely at the master's degree level) and offer them a part-time position. Maybe a revenue management grad student will enjoy this as well, and bonus points if you anonymize your data and make the project open source on GitHub. You could get major support there.
These aspiring data scientists need to know a basic stack and have some data skills too. The easiest could be Python and R-studio, but there are many ways to do this. It's not easy, but there is a reason data science is one of the fastest-growing technical fields, and it's about time it found its way into hostels. If you do make it open source, send us a link!
If you read this, kind of understood it, but it was still a little over your head, congratulations, you were learning about algorithms and data science. Just remember, revenue management is more than numbers; it is more than your bed rate. It is all about your tRev, your total revenue, and data science can really help you learn more about your customers.