The unpredicted results of both Brexit and the US election may have called data science into question but, says Wyndham, that isn't a reason to give up
On the day the US election results were announced Kelly McGuire, Vice President, Advanced Analytics, Wyndham Destination Network, had a call scheduled with her team. “They seemed a bit shaken,” she says, “but when I asked why I got an unexpected response”.
The advanced analytics team at Wyndham was upset because they saw the outcome as a failure of data science. After all, even the most respected, most accurate, pollsters had failed to predict the result. In a field that has only recently started to gain traction and acceptance, this was seen as a huge setback.
Data science absolutely does work but…is heavily dependent on how it is collected and the methodology used to analyse it
However, while McGuire is clear that there is an important lesson about data science to take from the US election, she is also clear that it isn’t that data science doesn’t work. Data science, she argues, absolutely does work but not only is it heavily dependent on how it is collected and the methodology used to analyse it, but is always subject to human intervention and bias.
Lessons learnt and looking ahead
Going forward into 2017, data analysts must, she argues, continue to ask the right questions. Where does the data come from? How was it collected? Is it a truly representation of the population?
“A lot more thought must go into the collection methodology and any potential biases that might arise, than even the analysis method and interpretation of the outcomes,” says McGuire who cites a favourite quote from the late British economist Ronald Coase that “if you torture the data long enough, it will confess anything”.
If you torture the data long enough, it will confess anything
If we take any lessons from the ‘surprising’ outcome of the US presidential election – or Brexit for that matter – it is to always question how the data is collected and be very cautious and responsible in the way we interpret the results.
Going forward, the turbulence that dominated in 2016 will likely continue into 2017. Against this backdrop of complexity and uncertainty, McGuire stresses that “hospitality organisations must continue to prioritise data-driven decision-making, and with that, perhaps more importantly, cross-functional collaboration”.
Because as all hotel brands know, in today’s competitive world of travel, the lines between functional areas are blurring, and the pace of business is accelerating. So, organisations need every perspective, backed up with data, to be able to act nimbly in the years ahead.
In 2017, Wyndham’s data & analytics priorities will include:
Continuing to invest in any technology that helps harness and drive insight from big data in new and creative ways
Pushing the team to innovate and drive results for the business
Reach out to counterparts across the organisation to understand the challenges they are facing, the decisions they are making and the insights they have, and see how her team’s knowledge and capability can help
Continuing to push the industry to invest in the right technology and resources to grow their data analytics capabilities.
To hear more from Wyndham’s Kelly McGuire and other big players and innovative upstarts including Accor, Delta, Hilton and Vacasa, join us at the Smart Data Summit North America, Atlanta, February 22-23
February 2017, Atlanta