Big Data, Big Analytics and Revenue Management
By Kelly McGuire Vice President, Advanced Analytics, Wyndham Destination Network | March 09, 2014
Revenue Management Data Really is Big Data
The most useful definition I have come across for big data is from Forrester: "When the volume, variety and velocity of data exceeds the organization's storage and processing capabilities for accurate and timely decision making".
I speak to many hoteliers who are not convinced that their organizations have big data. Let me lay out the revenue management problem for you, (typically solved using an analytical revenue management solution), so that I can demonstrate how revenue management has always been a big data problem. A typical hotel has the following input data:
- Detailed customer or market type segments (optimal for RM analytics): 60
- Different accommodation types: 12
- Historical dates (2 years of history): 730
- Future dates (1 year): 365
- Length of stay types: 8
- Snapshots stored for each occupancy date: 40
The combination of all of this input data for just one property is 252 million observations. If you then generate decisions based on this data and store those decisions, you will need to store approximately 10-20 gigabytes per property. For a hotel chain with 2,000-4,000 properties, that would equate to 20-80 terabytes of data. That is a lot of data. Note that this only includes only a subset of the information that revenue managers or a revenue management system might find valuable. It is likely that your organization is storing competitive rate information, STR data, forward looking demand data, and probably a few other data sources as well.
It is not just about volume though. Perhaps your revenue management system is handling the volume just fine right now. However, brand new data sources are cropping up all the time – data sources that could help to inform pricing decisions if they could be incorporated into the algorithms or available for ad-hoc analysis. Many of these new sources are in non-traditional formats – like unstructured text data from reviews, click-stream data from web interactions or geo-location data from mobile devices – which increases the variety of data. Some of this data changes so quickly, it is stale almost at the time it is created – like click-stream data or tweets. This is velocity.