Harnessing the Hotel's IQ

Mining the wealth of data that is available to hotels is both an art and a science. Analytics requires looking at historical and current intelligence combined with a dash of reading tea leaves to make smart decisions. With a combination of machine learning, predictive analytics and human intelligence, smart analytics requires nimble strategies and technology to empower staff to make service and operational improvements based on customer and business intelligence. In this executive roundtable, experts discuss what hotels can do to avoid mishandling and missing the benefits of available data.

In this article, HT talks to Michael Blake, CEO, Hospitality Technology Next Generation (HTNG) and Jim Walker, SVP of Global Revenue, Agilysys.

An often cited woe for businesses is being data rich, but insight poor. How can hotels overcome the paralysis of staggering amounts of data from multiple systems in order to make smart decisions about service and operations?

MICHAEL BLAKE: I think hotels spend a lot of time trying to get a purified data set and by the time they are done, it is obsolete. They also try to put the datasets in the hands of very few experts. These experts are generally overtasked and frustrated by not being able to fulfill requests. There are better tools now that enable you to manage large volume data sets. Data visualization toolsets also help to easily spot trends and make strategic change. These tools will be even more powerful when you can start operationalizing many of the insights in a daily operation. Perhaps manpower scheduling can be better optimized or your kitchen supply chain can be reconfigured or food waste reduced in the catering operations.

JIM WALKER: To get started, you need to know the end objective: what decisions must your data, and the resulting metrics, enable you to make? Consider starting with the information that you are lacking. Identify the gaps and start small. A small objective that does not interfere with guest service is usually a good starting point. 

To prioritize existing datasets, ask yourself, does the existing data feed your business goals? Beyond that, take the time to assess and prioritize the types of reports you want to generate and which datasets will feed those reports. What do you want the data to be able to tell you? How often do you want access to the data, and who will use it? 

One of the biggest challenges for hotel analytics is a silo mentality concerning data. What role can systems integration play in how hotels can fully leverage analytics? 

BLAKE: Currently data has been locked in silos of systems. HTNG workgroups are in place to help better unlock these data islands so handoffs between systems are easier and happen more seamlessly. Having vendors and hoteliers engaged with these workgroups enable them to set the direction and lead the industry as opposed to living with the outcome. 

WALKER: We have all heard that lack of data is the largest challenge facing operators. But it’s the lack of integrated data that’s the real challenge. The data we have access to might be abundant, but it is often disconnected and can divide an organization. 

Achieving the best results requires both careful planning and a complete understanding of the data sources. The more data sources you have, the more rigorous you must be in planning data integration, but the outcome is worth the time invested. Data integration projects, managed well, can give you the data you need to make better decisions faster and improve guest service as well as efficiency.

What are some common mistakes hotel companies make when dealing with data? 

BLAKE: Let me start out with no one is using data effectively in our industry. We spend a lot of time trying to understand guests and guest behavior yet we are surprised by bad trip advisor scores and stunned when we disservice a guest. I would say we have not distilled guest information to operationalize it and use it on a day-to-day basis. Some of it is the simple stuff, can you stack rank your guests in a given hotel on a given night? Some hotels may have over 80 VIPs but what 12 should the GM visit? The algorithm behind how you define the “best” guest may be what you need a data scientist or analytical resource for, but the end product of a stacked ranking is one thing a front desk person can and will respond to. You can operationalize how you may differentiate service to your top 10 guests or at a worst case make the front desk aware.

The harder stuff would be to try to figure out an individual price for a guest based on something other than past behavior so you are trying to understand a guest from a total value prospective. Current loyalty systems are not geared for the total contribution a guest can make to a hotel reputation or a current or future revenue basis.

What are the benefits of using data visualization to make information digestible for employees? 

WALKER: Successful operations require open, flexible communication. This is where data visualization can facilitate team collaboration; it helps tell your story. For example, dashboards that include alerts or other business health status indicators are ideal outputs. 

One of the best ways to convey performance against those objectives is through charts and graphs. Visual representation of the data imparts a sense of accountability and ownership because it clearly illustrates how well the company is performing, and what needs to be done to accomplish the objectives. It demonstrates that everyone has a part to play and helps inspire cooperation.  

How do security concerns affect the output of analytics? What measures can hotels take to make sure that data is secure while yielding value to brands?

BLAKE: One of the problems of democratizing data is that the wrong people will get the wrong data. I would say as long as you are not revealing PII or other secret sauce, allowing folks to use data and free the gatekeepers from the Sisyphean task of trying to satiate the masses. Security around data should also follow some common sense practices like knowing who is accessing the data and understanding why they need the data.

What can hotels do in order to be proactive rather than reactive with data and analytics?

WALKER: Proactive, and predictive, analytics should be the standard approach of information gathering and subsequent decision-making. Solidifying a proactive approach supports your efforts to promote more data-centric decision-making. This might mean proactively taking advantage of a sudden surge in spending or reducing pricing when catering sales declines, for example. More robust data analysis can show correlations between types of data in which a trend in one area consistently precedes the same trend in another area. In that case, the former can be used as a leading indicator for predictive analysis related to the latter.

There is still room for reactive analytics however. It can help explain what happened in a specific situation so you can correct it and move on. While it does not allow you to plan or anticipate certain results, it can help in certain situations.