Machine learning: the future of travel and hospitality

Machine learning is revolutionising the travel and hospitality industry with its ability to automate processes, engage customers and provide novel business insights.

The automation of revenue management alone means businesses are able to set prices and update rates automatically, freeing time for revenue managers to focus on tactics and strategy. Even more, by combining historic data with competitor data and booking trends, AI is able to accurately forecast demand, boosting hotel revenue in unprecedented ways.

Major hotel brands are seeing revenues increase by millions of dollars a year, with even small properties experiencing incremental sales lift by more than 15%. And this is just one of the many ways that machine learning can transform a business.

Three aspects of AI

How and why is machine learning having such a massive influence? There are three main areas where this new technology is adding real value to businesses. The first is in automating processes. Repetitive back-end procedures that would once have taken time and staff resources can now be completed automatically, in no time at all.

The second is in new forms of customer interaction. Using natural language processing (NLP), applications can be built which take over customer interactions such as booking requests, queries, complaints and other customer services. This frees up staff to focus on higher level tasks and provides a speedy, round-the-clock and consistent quality of customer service across the business.

Thirdly, given the vast amount of data that hotels produce on a daily basis, AI can mine into this and provide novel business insights, predictions and ways to maximise revenue through more personalised consumer offerings.

All three of these approaches fit perfectly with the increasingly digitalised, online and mobile nature of today’s travel and hospitality market.

Sciant — industry-specific expertise

At Sciant we have a proven track record of providing machine learning solutions for all three of the approaches outlined above. These include:

• Deploying smart machine learning into existing systems; we take your existing infrastructure and help identify areas ripe for automation.
• Developing sophisticated, sector-focused Natural Language Processing (NLP); with an array of uses from voice recognition to processing text, NLP can better assist staff with processes
• Ensuring seamless backend processes to check and match data; since data is a valuable commodity today, it needs to be relevant and clean to ensure automation is optimised. Here machine learning can help keep data clean.
• Optimising booking and reservation platforms for greater efficiency; a plethora of systems provides travel businesses with complex distribution and booking structures. Machine learning can consolidate and automate processes to reduce manual repetitive tasks.
• Creation of complex selection logic to process information. Machine learning can be limitless. We find ways to simplify processes even if it means developing sophisticated logic.
• Building predictive cache for systems handling huge quantities of data; the larger the business and number of customers, the bigger the data warehouse for processing information. We find ways to optimise the performance of these that help find commercial opportunities for the business.

And because Sciant specialises in hospitality and travel, we can apply these technologies to the specific systems and data sets within the industry with an in-depth knowledge of the requirements that businesses have and the challenges they face. This avoids potentially hundreds of wasted hours of industry-related knowledge transfer which a non-industry specific technology provider might typically require.

Case study — Hotel Res Bot

One example of how we used machine learning to add value and improve efficiency is with HERA (Hotel Email Reservation Assistant) by Hotel Res Bot. This is a system for automatically responding to hotel bookings via email.

HERA was designed to save time and free up staff from the repetitive task of replying to email reservations — up to 25% of all hotel bookings — by automating the process. Sciant provided the development support of the Natural Language Processing which allowed the system to read, understand and respond to natural language used by customers in emails, even handling multiple reservation requests within a single email.

Furthermore our deep industry knowledge allowed us to build seamless integrations with hotel systems such as property management systems and central reservation systems, enabling the software to check room rates and availability to make sure guests receive the best matching room and rate type.

Hotel Res Bot said of their collaboration with us: “Sciant’s hotel tech skills and knowledge was essential, especially the scientific element of helping hotels and solving problems as and when they arise. We’ve saved a lot of time because we have needed significantly less knowledge transfer which has resulted in more creative discussions on solutions and development. It has been a refreshing change to not have to educate developers on the ABCs of hotel tech.”

The approach — Successful implementation with Sciant

One of the reasons for the success of our projects is clearly-defined goals combined with excellent project management.

• We work with customers to identify the area that will most benefit from AI and from here we build an implementation plan.
• We set clearly-defined goals and the problems we want AI to solve.
• We provide full technical support in preparing data for AI implementation, providing cleaned and integrated data across all systems within the business.
• We help with planning how data is organised and accessed by AI, providing the optimum conditions for it to learn faster and more effectively.

Once the groundwork is laid we move towards full integration by implementing a small pilot project that provides an easily manageable stepping stone to larger integration. On the back of this we integrate AI incrementally into wider systems with ongoing support and feedback, with the ultimate goal of full, business-wide integration.

Challenges of AI implementation and the benefits of outsourcing

The two main hurdles for successful AI integration are data and staff. Data needs to be cleaned and integrated across all systems, not siloed in individual departments or locked away in legacy software. And the organisation needs experts trained specifically in AI and data science, not just a generic IT background.

This is why outsourcing AI integration to industry-expert technology providers like Sciant is so much more efficient, simple and cost-effective than trying to train or employ in-house staff and manage projects internally. It really does take the headache away.

Machine learning will change the future

Machine learning is already saving the travel and hospitality industry millions of dollars through automation, customer service, marketing and novel business insights, but the journey has only just begun. There are so many more processes that are ripe for automation, so many day-to-day improvements on AI’s language processing skills, and so much data that is just waiting to be analysed that the potential is almost endless.

Investing in machine learning is not just about adding value today, it’s about taking the first step on a journey into the future, a journey that could lead anywhere you want it to.