Change Region

You're on our English website. Change your region to see information for another location.

Why you should stop using CompSets to make pricing decisions

Originally posted on PaceRevenue.com by David Lopez Mateos, Chief Science Officer.

Using CompSets is fine for benchmarking but not when it comes to making pricing decisions. In an age of increasing online bookings, hotel revenue managers too often get caught up in maintaining parity with their CompSet. Many still operate using the old school revenue management approach, relying heavily on their Smith Travel Research Competitive Set analysis. As a tool and gauge, use the STR CompSet for a quick benchmark but don’t use it as a basis for a pricing or marketing strategy to individual consumers. Here's why:

“ Primary data-driven strategies are better at optimising revenue 80% of the time.”

CompSet prices are a particularly mysterious form of alchemy. They are the epitome of secondary (external) data and can be driven by myriad factors that will always be unknown to you - how many rooms a property has left to sell, how the property is being perceived in the market, how they have done so far this year, how the revenue manager is feeling that morning (to mention just a few). It’s arbitrary, at best. Those CompSets are being driven by irrelevant factors like employee emotions and outdated RM systems that are not fit for purpose.

Prudent and sensible revenue managers have no idea what ingredients are going into this sorcery. Is it a good idea to adjust your prices based on competitors' erroneously concocted prices? We argue that it's not.

We make the case that, in contrast to using the witchcraft of secondary data, your primary data (internal hotel data sources, such as transactions) is all that you need to plug in. Your internal data tells you exactly how much demand you have, how much you need, and at what prices demand is being realised. Ultimately, this is all you should need to know in order to maximise your revenue.

Pace doesn’t look at competition pricing - IDeaS and Duetto do. It’s a totally different approach and here is why we're right:

Clean vs noisy data

It's a statistical principle that the data that is the closest driver of a variable is the data that will always produce the cleanest measurement of that variable. This will, in turn, lead to better decision making. The further removed the data that you use is from the variable you actually want to measure, the worse your measurement will be. More factors add noise, which is not good for decision making.

What does this jargon mean in the context of hospitality? Well, in your case, it implies that you only really want to be using your own internal demand data. Ultimately, you want to measure demand at a given price (or all prices) and only then will you be able to find the price that maximises revenue. Your internal demand is not a proxy - it’s exactly what you need.

That’s another way of saying primary data matters alone and not secondary data. I’ll say it again for emphasis - your own data about your own demand is, statistically, the closest driver of what you are trying to understand. Your decision on pricing should be based on your own performance.

You rarely know entirely what the competition is putting into their pricing strategy. The reality is they are often making mistakes. You don’t want to be a follower, particularly if you are following your competitors off the proverbial cliff. Instead, be a market leader.

Results never lie

In the end, it boils down to what data is actually maximising a hotel’s revenue. What we’ve consistently seen when comparing hotels that exclusively use primary data algorithms versus those that exclusively use secondary data is that the former was superior at optimising revenue, particularly if they were capable of quickly adapting to market changes. To be more specific, primary data-driven strategies were better at optimising revenue 80% of the time.

In the chart below you can see the correlation between competitor’s rates and market demand for a typical property for one night in July 2020. You can see competitor rates have no correlation with demand whatsoever.

Pace-Revenue-correlation-between-competitor-rates-and-demand
Source: Pace Revenue

The chart below is the same plot for the whole month of July last year. You can see slightly weak correlations at the very low and very high end of the prices, but nonetheless, it vividly highlights that competitor prices are useless for gauging demand.

Pace-Revenue-correlation-between-competitor-prices-and-demand-over-one-month2

Source: Pace Revenue

Hospitality revenue management has been left behind. The cutting edge tools and technology have not been commercially available to hoteliers, until now. There have been major technological progress in other fields and it’s time to update things. But while you are doing so, please don’t use CompSets to make pricing decisions.


About Pace Revenue

We share a conviction that revenue management should be data-driven, not opinion-based. To achieve our mission, we have assembled an amazing team of data scientists, product experts, hard-core engineers and revenue managers. Learn how revenue management is changing by booking a demo with the Pace team.


Check out the other articles that make up our regular revenue management column.

Related reading