How international hotel giant IHG is using artificial intelligence to bring HR and recruitment benefits
You can't have your cake and eat it. This popular English phrase has been used in its contrary sense in recent times to explain everything from the UK’s desired outcome from Brexit to what millennials expect in the workplace. We all want our cake and to eat it. But in the world of hospitality, where indeed you would hope not only to have a cake but to revel in the sumptuous moment when you first eat it, this phrase is most apt when talking about today’s customer.
A report by one of the world’s largest hospitality businesses, InterContinental Hotels Group (IHG), launched at the World Economic Forum at Davos earlier this year, highlighted that today’s customers are uncompromising. They expect brands to deliver experiences that satisfy contradictory needs. In a landscape constantly changing through advance in technology, customers do not want either/or solutions. They want the best of both worlds.
To meet these expectations, organisations need to develop what IHG calls ambidextrous brand-building teams: teams of divergent thinkers with individual strengths and passions who can also work in an integrated manner to drive brand success. But for IHG, with more than 10 leading brands within the European market alone, all with distinct service offerings for different markets, getting the right people on such teams and delivering sustainable high performance and service experience is a complex and resource intensive process.
Given the scale of the European operation for hiring, training and appraising performance – we recognised that there were emerging opportunities for AI to play a part in helping to find the right team member who would best align and fit with each of the distinct brand personalities
To help simplify this, the $1.75 billion company, which franchises, leases, manages or owns nearly 5,200 hotels in just under 100 countries employing some 350,000 people, explored how artificial intelligence (AI) could help its European division innovate within the people space. Through a series of extended pilots across a number of hotel chains, it collaborated with UK based predictive analytics company Cognisess to pioneer a number of projects to pinpoint what impacts and benefits AI could bring within a traditional HR environment. Recruitment was an obvious first area to try as major benefits in accuracy, consistency and the removal of bias could be achieved – particularly given the scale of the task.
“The first and last impression guests have of our brands is very much engendered by our people. So, anything we can do to recruit, motivate and engage our people in delivering our brand promise is critical to our success,” explains Hazel Hogben, head of HR, hotel operations, Europe at IHG.
“Deploying the right people to deliver the true brand experience is a huge undertaking with our focus on increasing employee retention and improving performance in order to maximise customer satisfaction. Given the scale of the European operation for hiring, training and appraising performance – we recognised that there were emerging opportunities for AI to play a part in helping to find the right team member who would best align and fit with each of the distinct brand personalities.”
How's it done?
In a bid to firstly understand the human dimension of each distinctive hotel brand, the top performing customer-facing senior managers in each of the different brands were put through through assessments based on attributes such as cognition, emotional intelligence, personality types, behaviours and skills. This was correlated to their KPI performance data and customer data around service satisfaction and then cross referenced with the brand values of each different and distinctive hotel brands.
Analysing this data led to the development of a set of brand personas or ‘pen portraits’ matched to the typical manager in each brand, not only mirroring the brand’s unique essence and personality but highlighting where this was delivering a high-performance output and service experience. Algorithms for each brand were produced as individual profiles to be used in recruitment. These were translated into a series of assessments to test the potency of each candidate. Candidates were assessed against the 'pen portraits' via a series of gamification-based tools.
What's the impact?
On the back of the insight gained, IHG Europe’s HR and operations teams can profile internal and external candidates more accurately, increasing career mobility and also maximise the talent pool by ensuring the right talent can fit within the right brand proposition.
Recruiting managers in each hotel are able to quickly identify whether a candidate is a good fit for the job before an interview takes place. The assessments are unbiased and based on predictive performance, suitability and fit. This saves huge amount of times reviewing and interviewing multiple candidates and delivers better results.
"An immediate benefit of deploying this technology has been improving the candidate and recruiter experience. We’ve so far trialled this technology across a number of our Swiss hotels and had really positive feedback – candidates particularly love the gamification element and they generally find it fun, user-friendly and refreshing," says Hogben. “We’ve also had a strong response from our general managers who are finding it to be an incredibly useful, insightful and time-efficient tool in helping with the first stage of the interview process. As recruiters, they recognise the benefits of the technology eliminating personal or preconceived bias and this enables them to adopt the approach of hiring for personality fit and then training for skill levels. It contributes to making the initial stages of recruitment a more streamlined process and can also help to add value to existing roles, improve staff morale and retention levels. And, as we get a better understanding about the expectations and fit of the people we hire, we can help them to play to their strengths.”
AI and future leaders
From these initial findings, IHG recognised there were a number of other areas where there were some quick wins to be had, in particular its annual internal programme to fast-track potential future leaders through its employee development programme.
“Traditionally this entailed assessing a large cohort of colleagues from very diverse backgrounds and with very different levels of experience, making it hard to see why someone’s talent would or would not make a great future leader,” explains Hogen. “The process was previously run exclusively using several pre-screenings and assessment centres, making the process very resource intensive, and reliant on the personal judgement and expertise of the assessors. Using AI as part of our process was a perfect way of making our approach less resource intensive, less location independent (as most assessments were conducted online) and more transparent and objective.”
This was achieved by calibrating the platform by analysing performance data from previous years’ leadership cohorts based on IHG’s Winning Ways internal values and KPI metrics. This was then used to benchmark and rank the candidates with a model of what ‘best’ looks like by directing the machine learning algorithm with what to look for in the potential of new candidates.
This model, using comparative datasets, was deployed through online video assessments and facial and text analysis via our online platform. The algorithm mapped, ranked and predicted the expected performance of prospective candidates against criteria such as past performance indicators, internal values, personality and social cohesion. Consideration was also made to ensure the same amount of information applied to each person – thereby making it more transparent and inherently fairer.
Cognisess then measured this approach as a control measure against actual observer ratings within the physical assessment centres with the same cohort and demonstrated an accuracy of over 95% correct predictions on final choice of candidate and with equal accuracy for each of the metrics used to assess individual attributes. This proved that it was immensely quicker and cheaper to do this exercise in real-time on the platform compared to assessors spending hours and days on observing and rating the candidates manually and somewhat subjectively.
Based on the insights to date, Hogben expects to use AI in other areas of people and performance going forward. A mid-year and full year review element has recently been built into the platform to help capture and analyse employee performance. For the first time, IHG has been able to automate and speed up this process and more clearly identify talent potential, areas for development and intent to stay.
“We can see a lot of opportunities for further development of these tools. But for the time being we are just looking at piloting this approach across our Europe estate to better learn about our people and their potential.”
Where might AI play a significant role within the people space over the next one to two years? An interview with Dr Boris Altemeyer, chief science officer, Cognisess
“From today’s perspective with data availability and analytics capability, it seems strange to think that a 30 minute interview and a CV could ever provide nearly enough information to make a truly informed decision that is fair on the candidate and good for the organisation"
“Going forward complexity in assessing and deploying people will increase – not decrease. This is particularly pronounced in cases where, like the hospitality industry, dozens (if not hundreds) of people apply for a single role and the HR process needs to either be streamlined at the cost of transparency, fairness, and accuracy or it becomes very costly. Likewise, where re-organisation is necessary understanding which employees have the right talent to adapt or innovate within new structures and roles will be crucial.
"This suggests that predictive people analytics are the way forward in helping reduce complexity for humans and making large scale operations more cost effective while keeping it fair and transparent. However a common misconception is that using AI and predictive analytics will take the decision away from the HR practitioner and create ‘carbon copies’ of the status quo – far from it. This methodology will allow practitioners to confidently set very detailed expectations of the candidate or employee based on what has been shown to previously lead to actual good performance and use these factors to calibrate preferences which ensure that the candidates of today can more accurately fit for the requirements of the employment demands of tomorrow. In short, we believe AI can help people make better people decisions.
"Right now, we are already working with a number of leading companies to build new models and deployments for AI including looking at analysing and predicting team performance to help managers better assess how to construct and assess high-performing team dynamics. We are similarly looking at leadership attributes and qualities to understand where effective leadership can come from within an increasingly diverse and global workforce. This does not only relate to the obvious and important definitions of diversity in terms of background, gender, ethnicity or personality, it is also more crucially about the diversity of how people think and make sense of the world. Attributes like risk taking, decision making under pressure, or situational awareness are all aspects that will be increasingly needed for a workforce that is more connected than ever and one that will ultimately be expected to use their diversity to deliver organisational and societal success. These are all factors which can and will be made ‘tangible’ through the use of predictive people analytics.
"Equally, at a different end of the supply chain in the education space, AI can be used to better profile and understand talent from an early age to help increase social mobility and employability at crucial stages for young people going through higher and further education, apprenticeships and career planning."
What did you think about this content? Use the stars below to give it a rating out of five.