By Dr. Joshua Hayes, Allisha Miller, and Dr. David Hayes

For the third year in a row, Forbes[1] reported “Data Scientist” as the best job in America. The job pays well (median base salary of $110,000), and qualified data scientists are in high demand. Currently, Amazon is actively seeking to fill its 436 data scientist vacancies and Google is trying to hire another 135 data scientists.[2] Data science and data scientists are hot… and for good reason! The role of data scientists is to help businesses increase efficiency and profits by harnessing data. According to Amazon, “Data is the lifeblood of Amazon… Big data analytics is the magic wand for Amazon.

Hotels would like magic wands as well, and every hotel in the country could likely benefit from the services of a well-qualified data scientist. But the majority of hoteliers do not have the generous budgets of Amazon or Google. And they will not likely employ an exclusive on-site data scientist. Rather, in most cases, the hotel’s GM must serve as their property’s own big data analyst.

“Big data,” the identified power behind Amazon’s magic wand, is one of the most popular new managerial buzz words. But it can be hard to determine how revolutionary this change is because “big data” is only loosely defined. Many professionals use it with very different ideas in mind. Big data can simply refer to data that is large in size (usually over 1 Terabyte), or it can mean data large enough that it has to be stored across multiple locations. The past decade has seen the definition of big data change and grow to describe increasingly complex phenomena. For all practical purposes, however, big data means unprecedented amounts of data.

GMs know that their hotels already generate more data than previously imaginable, and the amount of available data gathered is increasing exponentially. In most cases, hotels generate far more data than they actually use. GMs only rarely need to capture entirely new data, so the much more crucial challenge is how GMs can best manage the data they already have. Keeping that in mind, the methods GMs use regarding their data becomes critically important. GMs could learn a lot from data scientists about how to approach data analysis methodically and efficiently.

To address even the most complicated analytical concerns for their businesses, data scientists are trained to assess an organization’s data set and ask the following crucial questions:

  1. What information is available?
  2. Is that information ready to be analyzed?
  3. Which type of analysis is appropriate?
  4. How can the results of the analysis guide profitable decision making?

These questions are not so different from those that GMs grapple with every day. However, while the work being done is not so different, it can sometimes seem like GMs and data scientists are living in different worlds and speaking different languages. As GMs try to learn about and incorporate the cutting-edge tools of data science, they often encounter loaded terms and concepts that confuse the process more than they clarify it. It doesn’t take long for newcomers to realize that seemingly common words (e.g. “cloud,” “java,” and “python”) seem to require a lot of specialized knowledge. Data scientists also love acronyms (e.g. R, SQL, CNN, NLP, TF-IDF, LCA, RDB just to name a few). The language divide can make the world of data science seem impenetrable. But you don’t need to master all the nuances of tech-talk to take advantage of the logic and insights of data scientists.

The truth is that data scientists do not live in a different world and they are not using magic wands. But those are myths that many data scientists are not eager to dispel. When GMs use an Excel spreadsheet they are already doing the same work as data scientists, perhaps just in a more simplified or limited capacity. Every day, and in all “normal” businesses around the world, key individuals use data to power their profit-oriented business decisions. Simply put, they turn data into profits. That work is not so distant from the surprisingly simple heart of data science: the smart application of data.

Data scientists are doing similar work to the average hotel general manager, just with rocket boosters. In fact, it may be more accurate to think of data scientists as racecar drivers instead of magicians. The top-level data scientists are truly remarkable individuals with honed skills, experience, and training. They’re capable of doing things that most people cannot, and they make it look easy. But most hotels can get along just fine without employing race car drivers. And GMs don’t need to be among the most elite data scientists in order to make meaningful, powerful, data-driven contributions to their hotels.

Most GMs are already comfortable managing large amounts of information. Given the vast amount of data their property’s already collect (from sources such as their CRS, PMS, POS, social media sites, proprietary- and chain-affiliated websites, and many more), it’s no surprise that today’s GMs find themselves awash in data. Increasingly, they recognize they must take a stronger role in properly managing that data. But without an idea of where to start, that can be overwhelming.

The specific data management challenges faced by hospitality mangers are growing increasingly complex[3], and this leads to continually changing training needs for GMs as time goes on.[4] There is no universal or simple solution to that overall problem. However, today’s GMs can begin to address their own data management challenges by applying the Obtain, Process, Analyze, and Apply (OPAA) method, which directly follows the 4-step big data management process used by the best data scientists:

OPAA Data Science Workflow for GMs

Step 1: Obtain data

This first step in the OPAA method directly addresses the data scientist’s question; “What information is available?” In this step, GMs must systematically inventory and assess all of the data sources currently available within their property and its various departments to see the range of data they can currently access.

The range will likely include primary sources of data such as the CRS, PMS, and POS, as well as secondary sources including in-house produced records (such as the Manager’s Daily), Smith Travel reports (i.e. STR reports), social media channel summaries, and even local weather reports. In this initial step, the GM’s goal is to determine what information (data) is currently available for assessment and which, if any, critical information is missing and must be obtained in the future.

Step 2: Process data

This step addresses the question, “Is that information ready to be analyzed?” The proper processing of data requires GMs ensure that systems for initial data entry and data records maintenance are in place and continually up-to-date.

For example, in a hotel’s sales department, pace/pick-up reports can provide valuable information to a variety of departmental managers, but these reports will only be accurate if confirmed reservations and sales contracts executed by sales staff are entered into the PMS accurately and in a timely manner. In this step, the emphasis is on maintaining the quality (rather than quantity) of a hotel’s data sources.

GMs addressing this step must also ensure their data is maintained in a form that allows for efficient analysis. Some example questions might be, “Are sales records maintained in Excel-ready documents or do they need to be altered from Word-based documents?” or “Do website tracking systems directly interface with the PMS or must key data be manually transferred?

The takeaway here is that information which can’t be analyzed is not helpful. If some data is critical, but not properly–formatted when initially recorded, GMs must change the status quo and introduce procedures to process critical information into an analysis-ready form.

Step 3: Analyze data

This key step addresses the question, “Which type of analysis is appropriate?” It is one thing to obtain and process accurate data, but quite another to determine how to best utilize that information to accurately assess business characteristics or market relationships.

When analyzing data, bigger is not inherently better. For example, a modern PMS system might have the ability to generate 50+ rooms-related informational reports per day, but it is not realistic to assume a front office manager will, on a daily basis, carefully review each of these reports. As a result, in this step, GMs work with department heads to determine how to compress all the data identified in Steps 1 and 2 into results that provide profitable decision making-related information.

It is important to recognize that even the most innovative, expansive, and flashy data science projects can often be boiled down into simple questions such as, “On average, is it faster to take Route A or Route B?” or, “Is the average score higher for group A or group B?

For hoteliers, a key question may be as simple as, “On average, will GOPPAR be higher if we charge Rate A or Rate B?

The takeaway here is that GMs don’t have to reinvent the analysis wheel. And they also don’t always need to harness the latest, most powerful AI techniques either. They can focus on identifying simple concepts (e.g. important numerical counts, averages, and percentages) that apply to critical areas (for example, room sales variances over time, revenue achieved from different reservation sources, or important operating cost ratios).

The most complex analysis isn’t helpful if it doesn’t connect to decisions that a business actually cares about. Similarly, the results from even a “simple” analysis, when applied to a key area, can mean huge benefits and increased profits.

Step 4: Apply data

This final step addresses the question, “How can the results of the analysis guide decision making?” While some GMs might seek to study the complex math and science behind the work of data scientists, the authors suggest that it is most often better for GMs to laser focus on what Amazon, Google, and others know well:

The ultimate goal of data science is to apply data in meaningful ways.

In this final step, GMs should provide concrete suggestions about how data summaries and analyses can be applied directly to make better, more profitable, business decisions. Data driven insights that help people make sense of the real world is at the heart of a data scientist’s work. This is because the information provided by data scientists is most valuable when it can be used to improve predictions of future outcomes, thus enhancing decision making.

For most GMs, key areas of data analysis will entail immediate, pragmatic problem-solving such as, “How much can we raise rates this coming weekend and still achieve our occupancy goals?” or, “Which online travel agency (OTA) provides our property with the best combination of rooms sold and net ADR?

In most cases, these kinds of questions, and many more, can be answered using the tools and information currently available to GMs, if they know how to Obtain, Process, Analyze, and Apply (OPAA) that data.

Now that the data scientist’s operating secrets are a bit clearer, GMs can better appreciate one fact that does distinguish data scientists from other analysts. Typical data scientists are not deterred when the available data does not immediately offer answers to the questions they are asking. Rather, the question, “What insights do we really want to get?” is the starting-point, and they will overcome whatever technological obstacles they have to in order to find their answers.

Once they know the challenge they want to tackle, they may even invent methods needed to navigate through all the other steps. They’ll find ways to capture information on tricky topics, often requiring meticulous (or creative) data-cleaning, and then find equally creative analyses to give insights into their initial questions. Successfully and efficiently tackling challenges in this way is an impressive task.

However, the hospitality industry at large has already identified the major insights that most GMs need to operate smoothly. Manufacturers have spent years creating a wide range of tools and incorporating those tools into hoteliers’ common operational systems. As a result, hotels may already possess many sophisticated analytical tools that their key managers just do not know how to utilize to their fullest.

GMs do not have to be an Indy 500 racecar driver to cover a lot of ground in regard to making data-driven business decisions; they just have to get behind the wheel of the data-oriented tools already available to them. That said, they will not be able to win their races until they effectively understand the ways that information is obtained, processed, analyzed, and applied to improve their decision making.

To be successful it is vital that GMs provide ongoing training and continual dialogue with department-level managers on how to apply the 4-step data science (OPAA) process in their areas of specialization. Some GMs may find that forming regularly scheduled meetings with their own in-house data management teams can help to empower those employees that can directly apply a hotel’s data-related discoveries. Others may find that specific training opportunities are needed to fully empower key personnel.

The most important take away is that, when done properly, GMs can go a long way towards filling the crucial role of data scientists. The result can greatly improve their own decision making, as well as that of their other key staff.

[1] https://www.forbes.com/sites/louiscolumbus/2018/01/29/data-scientist-is-the-best-job-in-america-according-glassdoors-2018-rankings/ [2] https://www.glassdoor.com/Jobs/Amazon-Data-Scientist-Jobs-EI_IE6036.0,6_KO7,21.htm [3] https://www.hotel-online.com/press_releases/release/data-data-and-more-data-the-most-valuable-currency-in-hospitality [4] https://www.hotel-online.com/press_releases/release/the-need-for-big-data-and-quantitative-skills-training-in-hospitality.

Note: This article originally published December 3, 2018 at www.hotel-online.com. Reuse by other media or news outlets or organizations are prohibited without permission. Personal use and sharing via social media tools is encouraged. All rights reserved by the authors.