The Emperor has no clothes: Why we need more data scientists in the hospitality industry
How Data Science Can Transform Conventional Hotel Hierarchies
Information Technology — Viewpoint by Henri Roelings
Oh, that's a good one! I think the problem with our industry is that we tend to fear data as the nemesis of the 'human touch' (whatever that means). Henri is right about hierarchical organizational structures. I live by a quote from Ed Catmull, who says: 'If there is more truth in the hallways than in meetings, you have a problem.' Apparently, this seems to be a recurring theme in our industry, and I think data scientists are feared because they can show that the Emperor actually has no clothes.
Let's expand on this topic:
Envision yourself aboard an aircraft: you pick up a magazine, settle into your seat, fasten your seatbelt, switch your phone to airplane mode, and get comfortable. Imagine a flight attendant approaching you with a question: 'Would you prefer the aircraft to be piloted by the captain or the co-pilot?' If your instinctive response is 'the captain, of course,' it may be prudent to reconsider. Surprisingly, statistics indicate that aircraft are more likely to crash when under the control of the most seasoned pilots. Malcolm Gladwell's book 'Outliers' delves into the disproportionately high incidence of plane crashes experienced by Korean Air in the 1990s. From 1988 to 1998, Korean Air's aircraft loss rate was nearly twenty times that of United Airlines, despite having access to well-maintained planes, excellent airports, and proficient crews. Analysis reveals that air disasters typically result from a sequence of minor errors rather than singular catastrophic events, and the National Transportation Safety Board identifies a typical pattern of seven consecutive human errors leading to a crash. The root cause of these issues may not lie in pilot competence but instead in cockpit communication dynamics. Cockpits require cooperative management, where tasks are shared and checked mutually to avoid or rectify mistakes.
The concept of the Power Distance Index (PDI), developed by Dutch psychologist Geert Hofstede, is pivotal here. PDI measures the extent to which less powerful members of organizations accept and expect power disparities. High PDI environments, like Korean Air's at the time, can stifle crucial communication, leading to unaddressed pilot errors. Recordings from black box data during Korea's problematic period often showed co-pilots hesitant to correct superiors due to rigid hierarchical norms, exacerbated by the Korean language's six levels of politeness. The shift in company policy in 2000, mandating English use in cockpit communication, marked a dramatic safety improvement, highlighting the impact of reducing PDI on operational success. This analogy extends to the hospitality industry, where hotels with lower PDI often perform better.
In our industry, strategic errors, like inappropriate pricing, can be swiftly corrected when staff feel empowered to look at data and communicate issues without fear of reprisal. This collaborative environment is crucial, as rigid hierarchies can prevent the timely correction of accumulating minor errors, potentially leading to significant losses. And this is where data (and data scientists) become incredibly precious. I hate talking about myself, but I think this is a good example: in 2010, I took over the management of a financially struggling hotel in Rome. By the end of the year, profits had increased by 25%, enabling the owner to expand. I won't name the hotel, but today, they have turned into a small chain, and most of the properties rank in the Top-5 on TripAdvisor for their respective cities. This turnaround was achieved not through significant investments or marketing strategies but by fostering an open, collaborative workplace culture. More importantly, it can be done by looking at data without any emotional bias. When I was wrong (and I was, often), I had a team that corrected me. I went the extra mile by relocating my office to be more accessible to staff, abolishing ties, and reducing formal barriers. This encouraged open communication, allowing staff to readily share insights and correct oversights, significantly enhancing operational efficiency (and revenue). I am not (and was not) a data scientist, but I am a data freak. And if you (and your team) can look at data clinically without being emotionally charged (a common issue in our ego-maniac industry), then you can really do great things.
The successful reduction of PDI at Korean Air provides a compelling argument for similar approaches in other industries. An environment where employees feel safe to base their views on data and data alone (and not fear hierarchical repercussions) is not just preferable; it is mandatory.