THANK YOU FOR SUBSCRIBING
Kelly A McGuire, VP, Analytics, Wyndham Exchange and Rentalsand Dexter E. Wood, Jr, SVP and Global Head Business & Investment Analysis of Hilton Worldwide.
Many hotel companies are embarking on analytics programs to improve the guest experience, maximize revenue and profits, optimize operations to control cost and increase the value of the guest relationship. Brands are hoping to drive value for owners through robust guest databases and advanced targeted marketing programs. Hospitality companies are looking to other industries such as retail, CPG and banking for inspiration on building analytics programs and analytics cultures.
This activity is, of course, very encouraging to those of us who have been evangelizing the value of analytics for years. However, many hotel companies are struggling to get these programs off of the ground, are not able to implement as fast as they would like, or are failing to see value from their efforts. The fault, as it turns out, is rarely the technology solution. The reason why these initiatives are so fraught with difficulty is that organizations fail to sufficiently plan for the data requirements that support them.
The hospitality industry is particularly impacted by this issue because data exists in myriad disparate systems, such as property management, central reservations and CRM systems. These systems tend to be either proprietary, or so highly customized that even data generated from systems from the same vendor could look very different. All of this creates a very complex data infrastructure that represents a huge barrier to analytical success.
There are many components to a data management strategy, much of which is under the complete purview of IT and the CIO. What we will focus on today is data governance-the portion of data management, which, while requiring a tight integration between business and IT, needs to be owned by the business if the reporting and analytics are to provide the value the business is hoping for.
Data governance refers to the process of gaining enterprise agreement on metric definitions, data sourcing, integrity, and security of the enterprise’s data. A data governance program includes a cross-functional governing body or council, a defined set of procedures, and a plan to execute those procedures. It also tends to be the most frequently overlooked or under resourced component of a data analytics project.
The importance of defining where the data comes from, how it is stored, who gets to access it and how it is used to calculate key metrics cannot be underestimated. Yet business users in different functions are often conflicted when it comes to agreeing on metric definitions, data sources and ownership. Those who have not experienced this process will have difficulty fully appreciating the amount of time and effort required. Nonetheless, the success or failure of any analytics project usually depends on a sound data governance program.
A Focused, Achievable Goal
Start by scoping out an achievable goal focused enough in scope to be clearly definable and also achievable in a reasonable amount of time. For example, Dexter’s team at Hilton wanted to improve performance analysis and reporting in advance of Hilton’s 2013 IPO. Their goal was to establish a “single source of the truth” for these critical activities. Because the data infrastructure at Hilton was complex and siloed, they first focused on ensuring consistency for the “up and out” reporting for the company’s senior executives, rather than trying to fix the data infrastructure across the entire organization.
Establishing a Common Business Language
The core of any data governance initiative is to develop what Dexter calls a “common business language”. Once the project goal is firmly established, the next step is to pull together a cross-functional team that can represent all the current and potential stakeholders that will access or use the data. Critically important is that the team members should be at a high enough level in the organization and have the authority to make and enforce decisions. IT needs to be represented along with the business, but it is important for the business to own the governance process. This team has three main goals: 1) to establish the source systems for the data, 2) to agree on the definition of key metrics, and 3) establish a business owner for each of the data sources to ensure sustainability in the program.
Reporting and analytics initiatives typically require application of business rules to raw data, whether summarizing (how many hotels in the Northeast), integrating (joining the data for two different brands) or deriving metrics (calculating occupancy). Without consistency in these rules, as they are applied across the organization, many different versions of the truth are created. The same information, in slightly different formats, can very easily end up sitting in the operations, finance, or revenue management department. Depending on what the different departments use the information for and how often it is updated, it can quickly get out of synch. This is why it is critical to establish a “single source of the truth.”
Consider the definition of hotel occupancy (rooms sold divided by hotel capacity). Hotel capacity is broadly defined as the number of rooms in the hotel. You would think this metric doesn’t change much. But what happens when a hotel goes under renovation and some rooms can’t be sold. Should they reduce hotel capacity? What about rooms that are out of order due to maintenance or comp rooms? As you can see, defining capacity can quickly get complicated. There may also be varying interpretations based on who is consuming the metric. This is why it is so crucial to have broad representation on the cross-functional team and to spend time discussing each metric in detail. The end product is the “common business language,” an agreed upon, well-defined and documented definition for every key metric used in analysis and reporting.
“The success or failure of any analytics project usually depends on a sound data governance program”
Data governance is not easy, but it does form the foundation for sustainable reporting and analytics. Putting the time in on data governance in advance saves time later. There are also other benefits to fostering this level of cross-functional cooperation as it becomes easier to synchronize decision making. Most importantly, without rules and business process around the data, it is easy for meetings to turn into debates about definitions and accuracy rather than interpretation and planning. When the entire organization is aligned, business users are more likely to trust the data. This means they will trust the results derived from that data, paving the way for success for analytics initiatives. Be disciplined, take the time, and you will see the rewards.