Generative AI as a Substitute for Social Graphs in Local Experience Discovery
A California-based startup is attempting to enhance its weekend planner site with a conversational AI interface. The goal is to offer personalized advice, mimicking the experience of chatting with a knowledgeable friend, according to co-founder Alex Ward. This new feature enables users to engage with curated local events, activities, and nightlife options through a back-and-forth text interface. For example, users could plan an itinerary featuring a scenic hike, a BBQ lunch, and an evening concert, or a cultural tour involving modern art museums, a painting class, and wine tasting.
Bigfoot, established in late 2022, initially received pre-seed funding and has built a database of over 120,000 events and activities across 160 cities. These events are accessible via category searches on its website. The latest addition is an AI chatbot which uses large language models (LLMs) from Anthropic, OpenAI, and Perplexity, marking a significant product relaunch aimed at making local discovery more engaging and user-friendly.
Building a Better Weekend Planner
The startup's platform peaked at around 30,000 monthly users for its first version, but financial constraints halted continuous marketing. According to Ward, the company's vision is to create a platform that small businesses can leverage to attract more local foot traffic. Securing seed funding is imperative to realize this vision, with the team believing that their blend of GenAI and over 50 curated data sources can offer more nuanced and relevant recommendations compared to general-purpose AI chatbot solutions like OpenAI's ChatGPT or Perplexity's AI search engine.
Challenges and Potential
However, the challenge remains stiff, especially from general-purpose AI tools which can also plan activity-filled weekends but usually end up offering predictable, tourist-heavy suggestions. Bigfoot's data sourcing, which spans mainstream platforms like Google Maps to specialized blogs, aims to provide unique and locally nuanced recommendations. For example, Littlefoot's suggestions for Barcelona included offbeat options like a hike at Parc del Mirador del Poble-Sec and an "alternative modernista tour."
Issues and Ongoing Development
Despite these efforts, the AI tool does have flaws. Users pointed out inconsistencies, such as the AI recommending indoor activities instead of the requested outdoor ones. This discrepancy often stems from relying too heavily on GenAI for event descriptions. The interface also needs improvements around plan presentation, as the current experience feels somewhat underdeveloped and requires better signposting and transition timings to ensure usability.
The startup's six-member team is working to refine its product, incorporating features like the ability to connect users' Google Maps or Spotify accounts to personalize recommendations further. Additionally, allowing users to upload notes and map pins to create sharable plans is in the pipeline. Bigfoot is actively seeking additional funding to support these developments.
In a competitive landscape where users typically turn to familiar platforms, the journey ahead for Bigfoot is challenging. However, their commitment to offering a curated and personalized local discovery experience may ultimately set them apart. The next steps include refining their AI's contextual understanding and improving the user interface to make the weekend planning experience seamless and engaging.