There is so much data that a hotel has or can have access to... but in most cases, there's no one who understands how to translate these data into actionable insights on a hotel's current and future business. A vast array of external and publically available data can influence a hotel's performance, its offering and allows the hotel to be better at what it delivers to its guests and its shareholder. Years of underutilized data are sitting in PMS, POS, CRM, and other systems. More than ever hotels need to justify investments and budgets. So will a data scientist be able to give these answers?
I fully agree that our industry is more and more driven by consumer behavior in a digital and data context. The hospitality sector has been slow to respond to the digital age which explains why other technology-driven companies have managed to dominate some of that space and changed the industry fundamentally. We need to move from a reactive to a proactive mode, not just to manage “in response” to new technology, but convert that influx of data into a strategic asset next to managing hotels and service experience we tend to focus on.
Most hospitality players lack that function and expertise and in our current times of great uncertainty and reset of consumer behavior that will accelerate during the pandemic, the need to use this data becomes paramount to pave the way for the recovery process.
Alone through a set of data is redundant. For it to yield knowledge, technology and human interaction are required. Further, even more critical are the skills of data interpretation into information, information into knowledge, and knowledge into business decisions. This is where a new hybrid function, the so-called data scientist, will start gaining credit.
The skillset these professionals will have to bring along will range from how to link the data between the various insular data pools that typically exist in hotel environments, interpret it to allow such professionals to identify trends and opportunities for optimization at any stage throughout the recovery window and well beyond.
The pace of change is increasing in the technological, socio-economic environment along with disruptions in guest behavior that by the time a data set becomes knowledge it is still relevant to guide business decisions into the future and ensuring that products can be scaled yet remain personalized. That new approach and function needs to interact with all stakeholders in a running hospitality business and may have a very good chance to develop from the rapid changes that are forming in recent times.
Before data scientists can to their job properly we need to find structured ways of collecting and consolidating our data. It's true that we have a lot of data floating around our industry but it still sits in many different places and even the ownership of the data is unclear.
Vendors often present flashy dashboards and BI tools but once implemented they fail to the deliver, not because the platforms aren't any good but because the underlying data is inconsistent and lacks integrity. I have worked with a number of data scientists and especially the ones from outside the industry are left frustrated.
To get the real power out of our data we need to have a strong data governance and then consolidate what we have in a consistent manner, perhaps through a data orchestration layer to start with.
In short, yes, but the problem hotels will have is the quality of the data and getting it into a repository in a meaningful, normalized, standardized way. Many legacy hotel systems are hell-bent on not allowing hotel staff to get at their own data, at least not at a reasonable cost. APIs are far and few between and those who start using them may find they are the first to do so and the APIs are partial and buggy. FTP file drops may be the only way to surface data. The traditional BI process is extract-transform-load-visualize, but the effort in the extract-transform piece can be 80% of the effort. Dashboards and data analysis are the easy part! Start early, invest properly, focus on customer data.
As a business we generate enormous amounts of data from multiple sources and want to and understand it not only to analyse current performance (i.e. business intelligence) but also to be able to accurately forecast.
In these testing times it is vital we understand not only sales patterns but also cost implications for future periods. We have always had this data, but having a data scientist / analyst team has given us the ability to understand business models, allowing the business to react faster to the ever changing world we find ourselves in.
Isn't it a shame: I'm having that ton of data available at my disposal – but, currently can't afford to further invest in talented people and high performance systems to give us more of these ever so helpful insights!
Never before in my career has it been more important to build actionable knowledge from data that will accelerate our recovery from the pandemic.
Yes, data-models and algorithms need to change to reflect the new situation – but this can be done by these talented people and powerful systems I (and many of my peers) cannot afford at this very moment in time.
But: I am convinced that it will get better again, our industry will recover and we will see these data driven projects starting up again.
Modern hoteliers realise that to win the hearts and minds of the digital customer, they have to effectively engage with those customers through all stages of the customer journey. This desire for continuous engagement in an ideal world manifests as a personal conversation via own-brand digital channels.
That's the desire, but the very real problem is stitching together sufficiently credible customer information from the vast array of customer data signals now available.
Customer data is held everywhere, and its mostly incomplete or just plain wrong, and in that format it is of low value and of little use to the hotelier. How often for example, does the front desk employee not use the customer information that's in front of them because they have learned the hard way that it cannot be trusted?
Enter the data scientist equipped with modern tools. Where have they come from? From other industry verticals most likely – Financial services, Telecom, and Retail have all been on this journey for some time, and together they have created a large market for modern tools and data services directed at building out customer-centric digital experiences using the large quantities of customer data, both self-generated and acquired, found across those industries.
Heading our way - Customer Data Platform (CDP) tools and Data Science practice that is central to CDP establishment and operation. Both of these will very soon become familiar terms.
Customer Data is about to get a whole lot better and much more useful. Get ready.
The need for both a person and for systems to make sense of diverse data has risen from a tactic to a strategy during these recent years. It isn't just the relatively low hanging fruit of analyzing data trapped within the PMS, CRS, CRM systems deployed at the property or in the cloud or within the corporate system, it is data about competitors' occupancy, pricing strategies and more that requires thoughtful outreach. I struggle with the term "data scientist," knowing that the "IT" person usually describes a job akin to a mechanic focused upon keeping the trains running, not pushing for new routes to serve expanded opportunities. Soon, there will be a need for three positions reporting to the CIO of larger groups: IT (the mechanic), RT (robot technologist), and DT (data technologist.) The DT role will become more important over time as the one to support growth of the organization.
The technology and data fragmentation in hospitality is a huge challenge and impediment to the post-COVID rebound the industry faces today. Data - guest data, comp set and market BI data - lives in multiple "data islands" that do not talk to each other: PMS, CRS, WBE, CRM, ORM, CMS, DMS, Social Media, and BI. Practically, no hotel company can boast a single view guest data that is being constantly refreshed and enriched with real-time, “live” data feeds from ALL touchpoints with the traveler and their Digital Customer Journey.
Quite often, different teams at the property or corporate use different sets of data in their day-to-day operations, creating a total "data integrity mess," which directly affects the property's guest acquisition and retention efforts. Ex. Most of the time, CRM data is not being utilized to engage and retain past guests, upsell amenities and services to on-property guests, and to create similar audiences in the property's new guest acquisition marketing efforts.
The goal here is very clear: bridge the guest data and technology silos and create an end-to-end real-time data platform, empowering hotels to acquire new guests, engage current guests, and retain past guests.
Here is the big question: Can the property hire a data scientist to do all that? The answer is no: there is a shortage of data scientists plus they are very expensive. The average salary for a data scientist in the U.S. is $154,000/year (Glassdoor). Expedia employs 365 data scientists at an average salary of $135,000; Facebook - over 500 of those at $152,000/year.Very few hotel companies can afford that.
Therefore, I believe it is up to the hotel tech vendor sector to carry the torch and help the industry overcome its technology deficiencies and create solutions to bridge the data silos in order to better serve the rising tide of tech-savvy travel consumers. In this respect, encouraging are the new type of Open API Marketplaces, offered by cloud PMS vendors like the Opera Cloud PMS with APIs to over 3,100 third-party tech solutions; StayNTouch PMS, CloudBeds and Protel PMS own marketplaces, as well as API Marketplaces like SiteMinder, HappiCloud, and IReconU.
The lack of hospitality-focused technology and data sciences education exacerbates the problem. Hoteliers are ill-equipped not only to deal with the fragmented data situation in-house but also to work with outside tech vendors providing next-gen AI-powered data solutions. How many hospitality schools today teach hospitality technology courses? Only a few. New York University's Tisch Center of Hospitality is one of the few to have a course on Hospitality Technology, which is a great start to educating future hoteliers on the importance of technology and data science in the industry.
Data scientists or “Data whisperers” will be a key part of successful hotels in the future. Data-driven decisions will become the norm and gut feeling and the “we have always done it that way” will disappear. As we are getting better with technology and the acceleration of adoption has happened because of COVID-19, we must start utilizing the data that is available to us both internally and also externally. There is so much publicly available free data that we can already access that could be an excellent starting point. So yes, this position will exist in the future and the early adopters will reap the rewards in these very challenging times.
This is a valuable position for all businesses that want to make the shift to data-driven decision making, but is certainly a critical position for any medium or large business. Quite often, though, the role of a data scientist is misunderstood. Data scientists are not "report writers" - they need a deep understanding of your entire business and must be able to interpret the data in all your disparate systems. They should be well-versed in database design, math, statistics, and programming - languages like python and R are critical, as is the ability to write complex SQL queries.
A good data scientist is comfortable working with both structured and unstructured data, can implement and operate ETL systems, and can build and use machine learning models based on guest behaviors to provide predictive analytics. As data currently sits in silos with no easy way to link it together, they also need to be able to properly cleanse, standardize, and link data.
Their goal should be to first get access to all of the data, get it clean, properly linked, and enriched with additional valuable data sources - including behavioral data like visits to websites, views of ads, clicks on email links, service requests, chats, and social interactions with your business. Only once that is in progress should your data science team start doing data analysis.
Properly implemented, you will have much greater insight into the key segments of your guests, how to interact with them, who you are at risk of losing, and what you should offer them next. This is the type of personalized service your guests are used to now, so ignore this at your own risk!
I think a Data Scientist will be essential at the hotel level. Unfortunately, this is going to be the hottest and priciest labor category in the market. So, maybe the brands will be able to afford a few of them and configure toolsets to make the insights actionable at the hotel level. Software solution companies will also employ data scientists so their solutions also remain competitive and actionable. So, over time the configuration and choice of the toolset will have baked in the “smarts” of these data scientists. Others on property are going to be taught the ways of these jedi data scientists and in the future, all of the hotel teams and staff will need to interpret and utilize data for the guest experience.
“Data Scientist” – That is perfect term to describe my job for the last two (2) months. You see the Hotel Development Company where I work is trading in three (3) legacy systems (Time & Attendance, Payroll & Human Resources & Employee Benefits) for a shiny new Human Capital Management (HCM) System.
My purpose for living recently has been manage the massive amount of data involved in a hospitality project like this – seven (7) years' worth of detailed employee data, in fact. As a “Data Scientist” my job is to be the facilitator between the Executive Team, the Operations folks, our Accounting & IT staff, our General Managers & end users and the firm making our shiny new HCM system. There are very high expectations on both sides and a real need to justify the investment and bring the project in on time and on budget.
Would I trade this experience and all that I have learned from it? No Way – not everyone can be a “Data Scientist”! Those who can gain knowledge and experience that few others dare to try.