Tripadvisor is looking to create a competitive advantage by using Snowflake technology as a platform to launch innovative data-led products for its customers.
Rahul Todkar, Head of Data and AI at Tripadvisor, explained to diginomica how the Snowflake AI Data Cloud for Travel & Hospitality is helping his company to unify enterprise information and analytics across product, marketing, customer insights, and finance operations:
We had 10,000-plus data pipelines and 30,000 to 40,000-plus data sets. Things were getting out of control. We had a massive data footprint, which we needed to streamline, organize, structure, and rationalize in a well-thought-out schema.
Todkar joined Tripadvisor in January 2023, previously holding senior leadership positions at LinkedIn and Charles Schwab. The decision to implement Snowflake predated his arrival. However, he says Tripadvisor has a tried-and-tested formula for selecting its partners:
For any technology, we look across the industry and explore the best solutions that fit our needs. We generally use a proof-of-concept process to figure out what solution can work in a defined use case and then expand it.
The travel specialist sought service redundancy and a way to protect its data from accidental deletion. He says Snowflake enables secure data sharing across multiple business units and geographies - and to partners externally within the wider travel ecosystem:
As a company, we're always big on data privacy, especially in the EU, because we have a strong presence in the European market. We looked at other options. But we tried Snowflake for redundancy, data protection and data privacy. It was a big success, and we loved it. We then started growing that footprint and leveraged Snowflake for our other use cases.
Todkar says his role allows him to take a broad view of potential vendors, partners and platforms in the market. He says modern businesses face a complex and growing data ecosystem. One way to deal with this complexity is to work with a minimal number of trusted partners across a large number of areas:
I'm a big fan of having fewer data partners that can do multiple jobs: 'If we have one solution, can we do multiple things with that partner?' The main concern is ensuring all our data is in one place, and the platform has features built-in for reliability, quality, monitoring, and maintenance upgrades.
Todkar says the Snowflake platform gives his business "massive consolidation". Snowflake has helped Tripadvisor streamline its operations and open new avenues for growth and partnerships. The result is less time on IT support concerns and more time on value-creation:
Before Snowflake, we had to do a lot of maintenance and software upgrades. There were so many things we had to do. Using Snowflake makes everything so much easier.
Finally, the Snowflake cloud technology is Tripadvisor's base for exploring emerging areas, such as generative AI. Todkar says the platform makes it much easier to bring new ideas into production and out to market quickly:
We now have this data foundation. So, for any application or new use case we want to launch, the time to realize the business benefits has shrunk with Snowflake.
Todkar says he's overseen multiple data transformation journeys during his career. The biggest challenges tend to be cultural. He says C-suite buy-in is crucial to the success of any data project, and that's something that's prioritized at Tripadvisor:
The tone is set by our CEO Matt Goldberg and our leadership team, where we've declared data is our competitive advantage. When you have that level of belief and strong executive buy-in for data, everything - the mission and vision around what you're trying to do - becomes easy from a cultural standpoint.
He says companies that execute their data visions operationally must ensure people across the business understand the benefits that will be delivered:
Go through it as a process, where that understanding - the realization of data as a top priority, what it means, and how you think of consolidation - is approached and then driven as a massive transmission project.
Todkar says the education process associated with a new cloud-based data platform takes time, but top-level support is critical to success, as is an ability to ensure technological challenges are kept in check. Once again, consolidation on a platform like Snowflake is key:
Technologically, I would say one very important thing is establishing certain providers as the foundational partners. If you talk to any engineer, there's a tendency to explore different solutions and options. But the more you can crystalize or standardize your strategy on fewer key partners, the better.
Tripadvisor continues to look for fresh ways to exploit Snowflake. Todkar is excited by the tech company's rapid development of new AI features. In particular, he points to the release of Cortex, a suite of AI features that use Large Language Models (LLMs) to help companies build generative AI applications:
We're eager to try these tools out. Cortex is certainly one we have on our mind to try and use more. Anytime you can democratize access to data, that is a massive unlock for everybody in the company. And so we're looking at that opportunity.
He says the company is also curious about leveraging features related to data quality monitoring because that area is "so important". The long-term aim is to use the Snowflake platform and a range of technologies to embrace emerging technology:
We're looking at an expanded set of features around how to prototype AI applications faster. There are a lot of areas where these models will be quite beneficial.
Tripadvisor has already launched generative AI use cases in production, including the firm's Trip Planning tool. The tool uses LLMs and generative AI to make personalized suggestions to travellers. The technology stack powering this tool draws on a range of systems and services and gives users personalized travel guidance:
