Artificial Intelligence: a functional fit in or a disciplined tool?

It is early days for the use of artificial intelligence (AI) and companies are still finding their way about where it best fits. For now, as the technology beds down, there are two possible approaches, writes Tom Bacon

Where AI should sit on an organisational chart remains something of a puzzle. But as more and more airlines, not to mention other travel suppliers, begin to use AI it is one that needs to be tackled. There are still many questions. For example, should data science be part of the IT department? Or should AI experts be dispersed across the organisation? Perhaps the approach depends on how AI is applied within a given organisation, or more precisely, whether AI is a ‘function’ or a ‘tool’.

AI as a function

If AI is a unique function that will likely continue to undergo dramatic change over time, travel suppliers will benefit from a centralised organisational structure that can recruit, train, and motivate employees with the requisite skills. As a specialised function, AI will learn and evolve; it will establish ‘best practices’; it will monitor cross-industry applications for possible internal adoption. In this scenario, AI becomes a distinct discipline on its own - increasingly sophisticated, more and more applicable to a broader set of business issues.

Chatbot technology, for example, could be applied to customer service and to loyalty and to HR and to e-merchandising. If AI is viewed as a unique function, like finance, marketing and HR, it will seek opportunities across the organisation and seek to maximise its value in each functional area. With this view, AI will be best organised as a sub-department within IT; like IT, AI would be a highly technical service provider to the rest of the organisation. 

This appears to be the thinking behind Air Canada’s AI Labs. Air Canada seeks to be a leader in AI, both in the airline industry and across other industries, and is attracting leading AI-trained data scientists.

Meanwhile Airbnb, a trailblazer in travel-related AI, has chosen a similar approach. AI resides organisationally separate from the other functional departments – but Airbnb assigns the data scientists in its IT organisation to cross-functional teams that enable AI-driven processes to address a variety of business objectives. The teams include representatives of each associated business or functional units involved with a given objective. 

United Airlines, too, has AI specialists in its IT organisation and ‘embeds’ them within functional units.

AI as a tool

Another view of AI is that of a tool, more like budgeting or planning than like finance or HR. This view anticipates that AI will evolve more along business lines than as a distinct discipline on its own. AI currently includes chatbots and self-driving cars, face recognition, recommendation engines and optimisation mechanisms – each business application will follow its own evolution; each needs to be more closely tied to the distinct business objective than to any particular ‘science’. Viewed as a tool, AI is held to the individual objectives – increased efficiency, more revenue, greater customer satisfaction – of the various departments.  Each department, in turn, needs to employ multiple strategies – some potentially leveraging AI, some potentially based on other less technical processes – to meet their objectives going forward. In addition, when viewed as a tool, the objective of AI is to help existing employees in a department meet their functional goals; this organisational solution thus emphasises ‘people first’ rather than a ‘people-replacement’ focus. When positioned within a business unit, there is the potential for greater transparency, greater trust, and, in the end, more openness to change.

In all likelihood, however, AI will end up as both a function and a tool. Perhaps, especially in the early days, it will help for AI to be managed separately from the other functions as it continues to be largely experimental. As it matures, or as a particular application becomes more established, the business functions will likely move to bring its application into their functional purview to better capitalise on its proven success.

Tom Bacon has been in the business for 25 years. When he isn’t penning his regular column for EyeforTravel, he is an industry consultant in revenue optimisation, and leads audit teams for airline commercial activities including revenue management, scheduling and fleet planning. Want to find out more? Email Tom or visit his website

Related Reads

comments powered by Disqus