Big Data has paved the way for understanding consumer context and can help you better monetize and increase yield or so all the headlines will have you believe.
But, the view from the sales and marketing department is often:

"How is this different from [insert-hype-term]?".
Instead of bringing hope, the mention of Big Data makes many director of sales and marketing professionals cringe.
Here are some practical tips from the bleeding edge for hoteliers interested in using Big Data to improve sales performance.
Don’t panic – You have always had to deal with Big Data
According to Wikipedia,

"Big Data usually includes data sets with sizes beyond the ability of commonly used software tools to capture, curate, manage, and process the data within a tolerable elapsed time."
Being unable to process data ‘within a tolerable elapsed time’ is the issue, not the size.
From that standpoint, Big Data only creates more headache when you can’t compile, analyze, interpret, and act on the data in a timely manner.
This isn’t a new problem. Management always had to deal with data overload ever since exporting data from the PMS, CRS, CRM, POS, etc. into Excel.
We all have many more reports than we know what to do with, so in essence, we’ve been swimming in the wake created by Big Data for a while.
Size doesn’t matter; intolerably elapsed time does.
"Housekeep" Big Data to gain efficiency
To shorten the data processing time, the data has to be pristine. Just like your hotel rooms need housekeepers, so too does your Big Data.
It’s easier to have dirty data than dirty bathrooms, but the dirtiness is harder to spot. Has Afghanistan ever been your number one country of guest residence?
You need a data housekeeper and housekeeping procedures to ensure that the data you have and the data being collected are clean and ready for use.
For inspiration, look at what your financial controller does to ensure that your company’s financial data are correct. The systems, codes, and processes that accurately account for all of the hotel’s operations show that data discipline can be achieved. The question isn’t if, but how.
Cleanliness is next to godliness, and your Big Data is only useful when it’s clean.
Segment first, then average, to understand the ‘Typical’
We like averages; the entire industry uses averages as benchmarks for performance. But averages can be misleading. This is because the word ‘average’ has two meanings:
- Adding several amounts and dividing the total by the number of amounts, or
- What is usual, typical, or representative.
The first point does not automatically mean the second point, so which point did someone mean when they showed you an average?
Take a traditional metric like Average Daily Rate (ADR). What percentage of rooms sold last month was actually sold within $1 of the ADR achieved?
If it’s less than 50%, then your ADR is a calculated value, not the representative rate a customer pays. And it would be dangerous to interpret otherwise.
To make ADR or any other averages more useful, segment the data so that the average reflects the typical. If you’re a manager, develop a Pavlovian response whenever someone says the word ‘average’; ask ‘For what sub-segment?’ This means that you’ll have many more averages, rather than, for instance, one ADR for the whole hotel.
Interpreting calculated averages as the ‘typical’ can kill a business; make sure the averages you see aren’t hiding how your business is truly performing.
Focus on outliers to prioritize targeting
Outliers are customers who are very different from your typical customers, this is is why understanding what is truly an ‘average’ is so important.
For example, the guest who spends three times her room rate on Food and Beverages is a positive outlier; the guest who books on Hotwire and doesn’t spend a single penny on ancillary services may be a negative outlier.
By using an outlier strategy, you can prioritize your target markets. You want to replace negative outliers with positive ones; in other words, get more people like your best customers and fewer people like your worst customers.
If you are evaluating your segments with one dimension, you can usually use a normal distribution curve to identify the outliers.
If you are evaluating your segments with two or more dimensions, you can use a strategy quadrant (sometimes referred to as a 2x2 strategy matrix) to do this.
The outcome of identifying your hotel’s outliers forms a go-to-market strategy, and the results are often surprising (Kazaks, really?) or difficult to accept (why not Singaporeans? We’ve always targeted them!).
Digital marketers are often tasked with getting more positive outliers, and this is a good responsibility for them. Because you can precisely target in digital media with relatively little cost, good digital marketers will get good returns on advertising spend.
Expect revenues of between 10-20x advertising spend if your digital marketer understands the profile of the outliers, which leads into the next point.
Map data together to add context
If you have a lot of data from different systems, you should try to put the different data feeds together to provide additional context to the segments you’re going to target.
The easiest way to do this is to have all the data snap to a time-series. This just means looking at data from different systems over the same time period, at a daily granularity if possible, to see if there are trends that would suggest a relationship (or in math terms, correlations).
For example, chart your daily website traffic data, daily TripAdvisor ranking, and daily direct bookings over the same period to see if the three lines have similar up and down patterns. If there appears to be a relationship, then you need to test out whether the appearance is random or causal (for instance, does higher traffic yield more direct bookings?).
The biggest mistake people make is to assume correlation means causality; or what appears to be similar patterns means the lines are related. If so, you could end up surmising that Facebook will lose 80% of its users by 2017, or Princeton will have no students by 2021.
See, even Ivy Leagus people can have issues interpreting Big Data.
In the same way you wouldn’t want an amateur to do your finances, you don’t want an amateur to deal with your Big Data, particularly if your strategy is data-driven.
Because language is inexact (like ‘average’), numbers are easily calculable, and mind-dazzling charts are easy to create with Excel, sometimes an impressive looking report is just that, impressive looking.
Having someone in the sales and marketing department properly trained in statistics, applied mathematics, or even finance will go a long way in helping you make Big Data useful.
Increasingly, winning the battle of getting heads on pillows requires data-driven marketers. You will need these types of marketers in addition to the sales team to stay competitive in the coming years.
Having your marketers lead the charge in mining Big Data for profit-generating insights can translate into huge gains in your hotel’s market share.
NB:Hotel data image via Shutterstock.