Market Simulation Deck

• 

  Market Simulation Deck - cover

Inspiration

Traditional market research, like surveys and focus groups, is often slow, expensive, and yields filtered responses. We were inspired by the idea of bypassing these limitations entirely. Instead of asking customers what they want, we wanted to create a tool that allows strategists to simply listen to them, uncovering the authentic, unfiltered needs that emerge from realistic, spontaneous conversation.

What it does

Market Simulation Deck is a virtual market laboratory. It transforms raw demographic and location data into realistic, data-driven customer personas using the Qloo API. It then deploys these AI agents into a simulated conversation about a specific topic (e.g., "what's missing in our neighborhood?"). Finally, a second AI acts as a market analyst, reading the entire conversation to produce a concise, strategic debrief that highlights latent needs, key pain points, and actionable business opportunities.

How we built it

This application was intentionally built as a self-contained tool within a single HTML file, using pure, "vanilla" HTML, CSS, and JavaScript with no external dependencies or frameworks. This ensures maximum portability and zero setup for demonstration. The AI interactions (for both the personas and the analyst) are designed to connect Qloo APIs to LLM APIs like Google Gemini and OpenAI, but run in a fully-scripted, high-fidelity "Demo Mode" out-of-the-box, allowing for a perfect showcase without needing live API keys.

Challenges we ran into

The primary challenge was managing a complex, stateful user interface within the constraints of a single file without a framework like React. Choreographing a seamless and perfectly timed guided demo tour that was both interactive and foolproof required careful event handling to prevent race conditions and logical dead-ends. Ensuring the application could perform a complete state reset for repeatable demos was another significant challenge that required meticulous refactoring.

Accomplishments that we're proud of

We are most proud of creating a flawless, self-explanatory guided demo tour that makes the application's value proposition immediately clear. Architecting a complex, multi-step application with asynchronous operations into a single, portable HTML file is a significant technical accomplishment. Finally, the end-to-end workflow - from abstract data input to a final, actionable business strategy report - is a powerful proof-of-concept we're proud to showcase.

What we learned

This project reinforced the power of a guided user experience. The application's value became exponentially clearer once we implemented the step-by-step tour, proving that how you present a tool is as important as what it does. We also learned the immense potential of using AI not just as a conversational agent, but as a built-in analyst to translate qualitative dialogue into structured, strategic insights.

What's next for Market Simulation Deck

The immediate next step is to move by integrating more LLM APIs. For scalability and to add more advanced features, the application would be migrated to a modern JavaScript framework like React or Svelte. Future enhancements include allowing a dynamic number of agents (simulating a focus group), enabling mid-simulation "injects" to test hypotheses in real-time, and expanding the final debrief into a more advanced, interactive visual dashboard.

Built With

• ai
• css3
• html5
• javascript
• qloo