4 practical strategies to turn AI experiments into real business value

Generative AI’s rocky road to adoption: Despite initial excitement, generative AI has faced challenges in delivering on its promise, with many projects failing to move beyond the proof-of-concept stage.

• Gartner predicts that at least 30% of generative AI projects will be abandoned after the proof-of-concept stage by the end of 2025.
• The technology has reached the “trough of disillusionment” in Gartner’s Hype Cycle for Emerging Technologies, indicating a significant shift from the previous year’s enthusiasm.
• Many tech CEOs who once spoke enthusiastically about AI tools are now apologizing for mentioning generative AI, recognizing the growing disillusionment among professionals.

Success stories amidst skepticism: Despite widespread challenges, some organizations are successfully implementing generative AI and reaping tangible benefits.

• Rahul Todkar, head of data and AI at Tripadvisor, believes in the real impact and benefits of generative AI technologies.
• Tripadvisor has successfully implemented generative AI in production services, benefiting both customers and the company’s bottom line.

Key strategies for generative AI success: Tripadvisor’s experience offers valuable insights into turning generative AI experiments into real business value.

1. Find a suitable challenge:
   • Tripadvisor identified trip planning as a common pain point for travelers.
   • The company reimagined the process to make it easy, delightful, and highly personalized using generative AI.
2. Use the right technology:
   • Tripadvisor focused on two areas: foundations and models.
   • The company deployed Snowflake AI Data Cloud for Travel and Hospitality to ensure data was in the right place.
   • They combined off-the-shelf commercial models (like OpenAI’s GPT series) with internally developed recommendation systems.
3. Turn use cases into revenue:

• • Tripadvisor’s personalized recommendations led to up to three times more user engagement.
  • Increased engagement resulted in longer site visits and more revenue.
  • The company leveraged its vast data resources, including user forums, to provide highly contextual recommendations.

4. Make long-term innovation plans:
   • Tripadvisor implemented a review summarization feature using generative AI, which has been in production for about six months.
   • The company is exploring interactive and multimodal experiences using its vast collection of user-submitted photos and property images.
   • Future plans include personalizing image recommendations based on user preferences and dietary restrictions.

Tangible results and milestones: Tripadvisor’s generative AI initiatives have shown promising outcomes.

• The Trip Planning Solution launched by Tripadvisor crossed the milestone of one million-plus trips created within nine months.
• The technology has significantly boosted customer experience and engagement.
• Personalized recommendations have led to increased user satisfaction and revenue growth.

The importance of data foundations: Tripadvisor’s success highlights the critical role of robust data infrastructure in generative AI implementations.

• Having all data in one place, using solutions like Snowflake, is crucial for moving from experimentation to production.
• A thoughtful approach combining commercial models with internal development work can create the best experiences for users.

Future prospects and ongoing innovation: Tripadvisor continues to explore new applications for generative AI to enhance user experiences.

• The company is working on leveraging its vast image database to create more personalized and contextual recommendations.
• Future developments may include tailoring image recommendations based on dietary preferences and other user-specific factors.

Analyzing deeper: The path from hype to value: Tripadvisor’s success with generative AI demonstrates that while the technology may have entered a period of disillusionment, organizations that approach implementation strategically can still derive significant value. The key lies in identifying relevant use cases, building strong data foundations, and focusing on tangible user benefits rather than getting caught up in the hype. As more companies follow this approach, we may see a shift from generalized excitement about AI to more targeted, value-driven applications across various industries.