Marketing Arcade
A 10-game interactive marketing arcade that turns my homepage from a static portfolio into a playable GTM experience.
Portfolios are static. Judgment isn't.
Most personal portfolios follow the same path — read a headline, scroll, leave. For the audience I care about (CMOs, heads of marketing, founders, GTM and PMM leaders) I wanted the first interaction to show how I think about marketing, not just describe it.
Ten games. One to two minutes each.
Embedded into the homepage hero, each game tests real marketing judgment across a distinct discipline — not textbook recall.
GTM Speedrun
Rapid-fire GTM prioritization and decision-making.
PMM vs Everyone
Cross-functional scenarios across Sales, Product, Leadership, Demand Gen, and CS.
The $1M CMO
Marketing budget allocation and portfolio decisions.
Fix the Funnel
Diagnose conversion, activation, retention, pipeline, and revenue leaks.
Position This!
Product positioning, differentiation, and value communication.
ICP Sniper
Market selection, segmentation, urgency, economics, and ICP judgment.
Marketing BS Detector
Spot weak strategy, vanity metrics, buzzwords, and misleading assumptions.
Battlecard Battle
Competitive positioning, objection handling, and sales enablement.
Kill the Jargon
Convert generic language into clear, outcome-oriented messaging.
Human vs Machine
Strategic judgment where AI recommendations conflict with business context.
Deterministic by design
Each game runs 8 questions against a 110-second timer, with randomized question and answer order. Scoring is weighted to reward the quality of the decision — diagnosis, prioritization, commercial thinking, positioning — rather than simple right or wrong. Players get an insight after every answer, navigate back and forth with answers persisting, and finish with a score out of 1,000 and a performance tier from GTM Explorer to GTM Architect. Replay serves a freshly randomized set.
- 8 questions per game, 110-second timer
- Randomized questions and answer order
- Weighted, decision-quality scoring
- Educational insight after each answer
- Back/forward navigation with persistence
- Score out of 1,000 + performance tier
- Replay with a new question set
- Player name capture, leaderboard-ready
Zero AI credits per play
No model is called while a visitor plays — no LLM inference, no token cost, no credits burned per game. Questions, answer logic, scoring, and insights are all predefined.
That keeps the experience fast, predictable, cheap to operate, and safe to expose publicly — the AI work happened in building it, not in running it.
Editorial and executive, with a bit of play
The hero splits 45% personal positioning / 55% arcade on desktop, stacking to introduction → arcade on mobile with touch controls and no horizontal overflow. Built on my emerald-and-ivory brand system with restrained muted-gold accents, clean card treatments, hover elevation, and a high-contrast scorecard — sophistication over a flashy gaming aesthetic.
Built hands-on across the full pipeline
Game logic, scoring, and content came together with Claude Code, Codex, and ChatGPT; Lovable hosts it and took it live; Supabase handles backend data. Scores persist through a single shared model — player name, game ID, final score, decision score, speed bonus, timestamp, and metadata — so all 10 games run off one structure rather than separate tables. The data model is already shaped to support a global leaderboard, per-game rankings, score history, cumulative XP, and shareable scorecards.
Where marketing depth meets product building
This sits at the intersection of Product Marketing, GTM, UX, and hands-on building — turning marketing frameworks into a working product, designing for a specific senior audience inside tight homepage real estate, and connecting a frontend experience to a real backend. Strategic thinking and execution in one artifact.
A zero-AI-credit marketing arcade — 10 senior-level games across PMM, GTM, positioning, demand gen, competitive strategy, and AI judgment — embedded into my homepage for a 1–2 minute interactive experience. Built with Claude Code, Codex, ChatGPT, Lovable, and Supabase.