How to avoid “big data paralysis” - A Guide for Hoteliers


We have spent a good deal of time at the Analytic Hospitality Executive advocating for the value of big data for hospitality. Just a few months ago, for example, I wrote a two part series on how Big Data was a “big opportunity” for hotels and casinos. Our goal at this blog is to help you understand opportunities to leverage data and analytics to move your business forward. Big data and big analytics, and the technology to take advantage of them in particular is a complex and fast moving topic. New opportunities constantly present themselves. It is difficult to sort through what will be sustainable and what is a passing fad. It can be confusing, risky and uncertain. It is difficult to justify investment today when the game may be changed completely by tomorrow.

It is just this challenge that I want to address in this blog. With all the highly publicized opportunities in big data and the ever evolving technology landscape, most hospitality and travel organizations are proceeding with caution when it comes to this area – and with good reason. These are expensive investments with many moving pieces.

There has to be a balance, however. Hospitality companies need a solid data and analytics program to support their overall business strategy, and to stay competitive. This strategy should be carefully constructed in light of the business requirements and organizational goals. However, none of these initiatives have to be perfect right out of the gate. There is too much potential in the data to wait for the perfect data warehouse or the most robust analytics package with the most innovative real-time collection and delivery. This isn’t rocket science or brain surgery. Something is better than nothing. Directional guidance can be highly valuable, even if it doesn’t point you to the optimal solution.

I was reminded of this years ago when I was doing a consulting engagement at a casino buffet restaurant. Casino marketing was running a two-for-one promotion that was creating long waits for tables. So much so that the customer satisfaction scores that the managers were bonused on were suffering. We were called in to build some capacity or revenue management strategies to reduce wait times and increase throughput. In our initial meetings we brainstormed the idea of running an optimal table mix analysis – an optimization algorithm that matches table sizes to the party size distribution, reducing the number of empty seats. We needed to collect and analyze party size distribution, run the optimization algorithm and then scenario test to make sure the mix would stand up against the variability in the distribution.   When we returned a few weeks later to review initial findings, the manager pulled me aside. He told me he liked the table mix idea, figured he had too many four tops on the floor, and went ahead and replaced them with a bunch of tables of two. He told me he was doing about 36 more covers per hour, and that he felt like satisfaction was increasing. The actual analysis revealed that he was off by a few tables here or there, but the point was that he was able to take advantage of the opportunity to increase throughput and reduce wait times before the “perfect” answer came, just by using data gained by his own observation of the operations.

Before you bite off big data, you are better off working with what you have, driving value and using that to set the path forward.

Organizations need to think through how they can use their existing data to drive value today as they configure and grow the database. Testing and learning on what you have will help to inform future data acquisition and discovery plans, and help to prioritize future actions. You may discover that the data point you thought was crucial is just not that important. You will certainly uncover new sources or directions for the analysis and future data collection.

One of the initiatives I have been hearing a lot about through my travels is the opportunity to use free wifi sign up to do location based marketing. Providing this service to guests or patrons, in exchange for being able to do push marketing does provide an opportunity to interact, and potentially drive revenue. Retail, airports and casino integrated resorts seem most interested in this right now. At first my reaction was “well, what do you really know about that mobile user anyway? How useful can this limited data really be?” As I thought it through, I realized that the limited data actually represents a huge opportunity to get creative, to test and learn and to execute against a very contained problem, which has huge implications when more information is added. There is no point in waiting around for the profile to be complete and accurate – you can only get so much information from any one patron.  Taking each of these interactions together as a whole, however, using the opportunity to quickly test and learn, provides a big opportunity to drive value.

A recent study from our partners at the Cornell Center for Hospitality Research, “The Mobile Revolution is Here, Are You Ready?” conducted by Heather Linton and Rob Kwortnik, investigated traveler preferences for interacting with hotel companies through their smartphones. They found that most guests prefer to use their smart phones to automate routine tasks like checking in, ordering room service or contacting staff. If there is a problem, they prefer to talk to a person. Hotels are taking on many technology driven initiatives to reduce friction through the guest journey, like mobile check-in, tablet ordering for room service, and mobile phones to unlock doors. Many of these initiatives are operationally focused, and organizations do not appear to be thinking through what further insights could be gained from the data collected from them. Are there hidden opportunities for insight? The data from any of these service oriented initiatives, limited though it may be, could potentially add value to a predictive analysis, provide insights about the behavior of guest segments, and help you design service processes that increase guest engagement and drive guest value.

Think creatively about what you would do if you only had an identifying number (mobile number) and knew the location of the person holding that device at this moment. The mobile number itself provides a clue to where the guest is from – but that’s really about it. What kinds of offers or promotions could you send to them, keeping in mind that would be able to collect what they responded to and know whether that same mobile number returned to your property?   If they were having dinner in the upscale restaurant, maybe promote an after dinner cocktail in your lounge. Spent more than three minutes in front of the retail shop, a coupon for 20% off their purchase, good only for an hour from now. This is basic, there are probably some clever offers you can build that will get the wifi user to reveal some additional useful information about themselves that could be extended to others that exhibit the same behavior.   You will certainly not get it right every time, or even most of the time, but you can learn over time based on the redemption of the offers across the customer base.

The wifi vision involves some technology investment, and it is a big data problem because it involves real time decisioning, but I use it as an example of how even a small amount of data about guest behavior can provide valuable insight. A next step could be to tie back to guest profiles, so their location and promotional response enriches what you know about them, and their profile provides opportunities for more targeted promotions.

Many hotel companies are making investments in building a robust guest database to support more targeted marketing and facilitate personalization initiatives. You should make an effort to continue to gather more detailed information about your guests to enrich your guest database, but while you are doing that, why not work with a small select set of data to test out some options? Pick a dataset that you are confident is clean and credible, and work with the team to figure out what insight can be gained from it. Can it demonstrate the value of some more advanced segmentation analysis? Can you use it to test out the attractiveness of certain channels or certain promotions?

Given how competitive the environment is becoming and how fast technology is moving, hotels cannot afford to wait until everything is perfect before moving on to the next step. Applying advanced analytics to the data you have today can provide enough insight and value to justify moving forward with the data you want to collect for tomorrow. Avoid getting stuck on the idea that you must have robust big data sets before you can take any action.


About Author

Kelly McGuire

Comments are closed.

Back to Top