Discover Personalized AI Travel Expeditions

Experiment

We tested the hypothesis that travel influencers would engage with Wndrtrip if given a personalized AI-powered itinerary preview. We also tested whether a simple interactive experience could increase clicks and conversion to join the Wndrtrip influencer program.

Key Takeaways

• • The integration of a personalized AI-powered itinerary preview for travel influencers proved to be a compelling engagement strategy, highlighting the value of customization in modern digital marketing.
• • Incorporating an interactive user experience, like a catchy travel jingle, successfully enhanced the emotional appeal of the brand, especially among target demographics like Millennials and Gen Z.
• • Utilizing v0.dev as an AI app builder demonstrated its strength in facilitating rapid, iterative development, which can significantly reduce time-to-market for tech-centered projects.
• • The connection between v0.dev, Hubspot, and Supabase allowed for streamlined lead generation and data capture, providing valuable insights into traveler preferences and influencer interests.
• • Challenges like generic AI responses and limited styling options underscore the need for ongoing prompt refinement and manual tweaks in AI-powered design, indicating areas for further enhancement in future iterations.

App Builders Used

We used v0.dev as our AI app builder for its ability to quickly generate UI with AI features, and we connected it to Hubspot and Supabase. Hubspot is used to manage lead generations, while Supabase will give us insights on travelers’ preferences.

Prompting Strategies

We used ChatGpt to create structured prompting to define both the UI and functionality. Example prompt: “Create an onboarding form for travel influencers that asks about their content style, destination, and brand preferences.” We refined the prompt a few times to get the tone right and make sure the outputs felt aligned with the Wndrtrip brand. We also had to split the prompt into simpler parts so GPT could incrementally build exactly what we want.

Project Outcomes

What Worked Well

v0.dev allowed us to build and iterate quickly with UI and AI model integration. The AI-generated outputs felt on-brand and engaging. Connecting to Google Forms for data capture was simple and effective for our MVP.

What Didn't Work Well

Some GPT responses were too generic until we fine-tuned the prompts. Styling in v0.dev was somewhat limited without manual adjustments.

Surprising Findings

We were surprised by how effectively v0.dev could generate both frontend UI and backend functionality with minimal effort. Its a very powerful tool.

Moving Forward

It let us move beyond static landing pages and actually test engagement with our core product idea of AI-powered influencer campaign matching. It also gave us real user data to analyze preferences and interest levels, all within a few minutes.

Full Project Site