Skift Take
AI is only as good as the data it comes from and the human ingenuity that powers it. Now is not the time to back down or sit back. The travel industry has few better opportunities to engage, learn, and invest in data and AI. Teams who are in for the long haul will benefit from a head start and can drive innovation in the industry.
Vivek Bogaraju
Like any emerging technology, AI is not a foregone conclusion; realizing the full potential of AI in travel will require bold action within companies and collaboration across them.
Coming off of the inaugural Skift Data & AI Summit held in New York City earlier this month, the entire travel industry is hopeful about how it will embrace AI. It was great to see a common theme emerge throughout the day, not just on stage but in conversations among attendees.
“AI is only as good as the data that creates it” was a common theme at the summit. Clearly, this statement requires one important clarification: AI is only as good as the data that creates it, and the human ingenuity that powers it.
We must engage with AI, learn, and make an effort to unlearn. How the travel industry embraces AI and derives its value will depend on our individual and collective beliefs.
Below are five questions to consider as you navigate your travel data and AI efforts: The purpose of these questions and the recommendations that underpin them is not to predict the future of AI, but to provide a realistic view of the opportunities for data and AI in the travel industry.
1. What is the right approach to get started?
Commit to continuing learning
Although advances in AI are announced almost daily, the reality is that we are still at the beginning of the AI ​​era.
We don't know what we don't know. We must learn, experiment, and accept that there will be setbacks, failures, and mistakes. Our practices and assumptions will evolve. This is especially true for how we evaluate and resource our AI initiatives.
Plan ahead and adapt to multimodal and multimodel AI
We already know that AI models are adept at switching between text, voice, and vision as needed to remain effective and relevant – they are multimodal – and it’s possible that different AI models may be better at certain tasks than others.
We need to be cognizant of this multi-modal, multi-model approach in the projects we develop. User behavior can change over time, so we need to be flexible about how users want to interact with us. As our AI initiatives evolve, we may need to switch underlying AI models and technologies to improve results and relevance.
Source: Skift/Vivek Bhogaraju
Evaluating competing priorities and investments
Competing initiatives are often evaluated based on the value they create for the business, and with emerging technologies like AI, two additional factors play a key role: ease of execution and end-user acceptance.
Depending on the AI ​​use case, the evaluation criteria should be expanded.
Here are some examples: What are the ethical considerations regarding the use of this technology? If we build it, how do we ensure that our current biases, preconceptions, and agendas don't influence the new technology? Once we build it, how do we monitor it to make sure it's serving the purpose it was designed for?
Accept that changing goalposts are part of the journey
We must recognize that as technology evolves, initial assumptions change. Prescribed guidelines expand, new regulations are introduced, and even desired outcomes change. How do we respond to shifting benchmarks? An evolutionary approach that embraces a test-and-learn mindset may be best for helping teams and organizations navigate AI maturity over the next few years.
2. Are you looking at the data correctly?
Reconsider data sources, ethics, and data acquisition methods
Data is a prerequisite for AI. It's essential. This is especially true for travel. There's a lot of attention right now on the availability and abundance of data for AI, and we're already seeing data licensing deals from AI companies building the foundational models.
The travel industry is filled with methods for sourcing publicly available data from the internet. These methods are cumbersome and often produce inaccurate, inconsistent and unreliable results. Most importantly, they are unethical unless data providers choose to be part of the process. It's time to change these methods.
Source: Skift/Vivek Bhogaraju
Ensure data sources are sustainable and reliable
Now is the time to rethink data sources and focus on data acquired by developing integrations and using open APIs. This will ensure reliable and sustainable data streams. Let's rethink how we handle data and create commercial incentives for data providers to build a robust infrastructure for data sharing.
Integrating data across all subsectors of the travel industry
Rethinking infrastructure also presents a unique opportunity to address the lack of data availability across different travel subsectors, including activities, air, car, cruise, hotel, rail and short-term rentals.
Today, data is siloed in legacy platforms and travelers often have to complete bookings in multiple places to plan their trip. But it doesn't have to be that way. It's time to rethink how data is shared across these subsectors and build something better that simplifies the traveler experience.
Balancing innovation with commercial interests
When it comes to making data available for emerging technologies, there will always be healthy debates about the clash between innovation and commercial interests.
Innovation is fueled by the free flow of information, data and open source technologies. Innovation also requires funding and support from investors, who are often driven by profit, and proprietary technologies may be best suited for that purpose.
Once you’ve broken down your current data silos, you need to be careful not to replace them with new, paid data silos.
3. What makes an organization AI-friendly?
Find the right leaders and give them the space to perform
Chief Data Officers have very short tenures – 30 months on average, according to HBR – and they often have extensive technical knowledge but no travel industry experience – both are necessary to be successful.
Once adopted, they are time-constrained to deliver results, and organizations cut their losses when results aren't immediately evident or require more restructuring than originally thought. Data and AI initiatives and investments are not short-term endeavors.
Define the right organizational structure for data and AI
Everyone wants to implement AI initiatives, but not many people are thinking about organizational design.
Organizations need to find a home that gives their data and AI teams a holistic view and influence without getting caught up in turf wars.
We need to think hard about how these teams work together and ultimately hold these teams accountable.
Choose between Standalone and Embedded Teams
Essentially, organizations must decide whether to establish a separate data and AI team that interfaces with other departments, or to have team members across the organization so that every function can be infused with data and AI.
The latter sounds better than the former because it is, and it's also much harder to execute.
Today, there is no one-size-fits-all location for an AI team: each organization will address this based on its own unique characteristics.
Set expectations and establish handovers
Regardless of how your organization defines the appropriate place for data and AI, you need to know where one function ends and the other begins. Handoffs and clear expectations between teams are key.
Many teams across the organization will benefit from data and AI successes: analytics, revenue management, pricing, sales, loyalty, distribution, and more. But to make this happen, they will need to commit their own resources and align priorities.
4. What is the right mindset for travel to unlock the full potential of AI?
Travel is cyclical and seasonal
The travel industry is affected by weather, viruses, wars, etc. This is the nature of the travel industry.
This means there is a need for urgency and efficiency in implementing data and AI effectively. Travel companies often struggle to focus and resource on long-term strategic initiatives while adapting to changing market conditions.
Fighting the long game
Successful data and AI projects take time, iteration, and most importantly, persistence — and as the underlying technologies mature, your data and AI efforts will need to evolve.
The winners aren't necessarily those who got started early or those who jumped on the trend later.
The winners in data and AI will be those who are bold and passionate about their customers, who are willing to learn and unlearn, who are flexible in their tactics, and who persevere. If you're looking for something more substantial than chatbots in the long run, temper your expectations.
5. Why now is the best time to invest in AI
Don't let skeptics shape the narrative
There will always be naysayers in the travel industry who point out the industry's slow adoption of technology, let alone embracing emerging technologies like AI. AI readiness and maturity in the travel industry is the worst compared to other industries, yet the travel industry represents the greatest opportunity for AI. Our history doesn't have to define the future.
Overcoming industry-specific challenges
The travel industry has actually come a long way in adopting new technologies – just think about the rate at which consumer preferences, search and e-commerce are changing.
Scaling and adopting technology across the travel industry is difficult for a variety of reasons, including the global nature of the industry, the need to be hyper-locally relevant and useful to travelers, small technology providers, fragmented ownership structures, and a long tail of customers in each subsector.
Standing in this special moment
AI presents an opportunity to reshape how the travel industry derives insights from data, curates recommendations, and automates actions. This can only happen once at the beginning of a new technology wave. We must seize this opportunity to rewrite the rules and usher in a new era of innovation and value for all stakeholders in travel. Let's get started!
Vivek Bhogaraju is an Advisory Partner for Data and AI at Skift. Vivek partnered with the Skift team to host the inaugural Skift Data + AI Summit earlier this month. He has built his career at the intersection of travel, data and technology through recent stints at Expedia Group, IDeaS – A SAS Company, Lighthouse and Oberoi Hotels & Resorts.
