Maps Cost Estimator
A closed-loop growth experiment demonstrating full-stack instrumentation, AI-powered intent capture, and seamless lead generation.
Application Overview
Interactive Calculator
An AI-powered smart-fill calculator that estimates Google Maps Platform costs based on natural language app descriptions. Features real-time UI updates, subscription tiers, and pro-rata calculations.
MCP Server
A Model Context Protocol server exposing the pricing calculator as a tool to AI agents. It allows stateless, reliable integration for clients like Claude, Antigravity, and Cursor.
Analytics Dashboard
Real-time Data Studio (Looker) dashboard embedded natively. Visualizes high-volume BigQuery data captured from user interactions, slider selections, and quiz responses.
The Closed-Loop Growth Experiment
This application was built as a comprehensive growth experiment. It transforms the static pricing page into an interactive tool. We actively capture user intent, segments audiences, and provide actionable data for marketing and sales pipelines.
Acquisition & Engagement
Instead of overwhelming users with a table of 40+ SKUs and volume tiers, the UI uses Gemini AI to translate simple app descriptions into complete, configured API usage estimates. This lowers the barrier to entry and builds immediate trust.
Data Capture & A/B Testing
We use Firebase Remote Config to dynamically A/B test an interactive Quiz Modal against standard "Get Started" buttons. The quiz captures precise prospect data: their specific role, project goals, estimated budget, and the exact value of the quoted configuration.
Segmentation & Lead Gen
Lead data flows securely into Firestore and BigQuery. Users are automatically segmented based on their calculator interaction (e.g., "Developers" vs "Product Owners"). High-value leads can be instantly routed to sales teams or specialized email sequences based on their calculated quote value and user role.
Analytics & Optimization
The embedded Data Studio dashboard provides a live view of the entire funnel. By tracking everything from granular slider adjustments to quiz drop-offs and final conversion rates, we enable continuous optimization of both the the UI experience and segment marketing.
MCP & Developer Discovery
By exposing the pricing engine via the Model Context Protocol (MCP), the tool moves beyond a browser tab and into developer workflows. This enables discovery through MCP and Skill directories, allowing AI agents to pull accurate pricing data directly during the planning phase.
Systems Architecture
Backend & Data: Firebase App Hosting, Firestore, Remote Config, Google Analytics 4
Data Pipeline: GA4 Export → BigQuery → Data Studio (Looker) Dashboard
AI & Integration: Gemini AI SDK, Antigravity, Model Context Protocol (MCP)