Homework : Suno Journey A/B Test Experience

# Homework : Suno Journey A/B Test Experience :::info - User Journey (從開始用到現在的體驗) - Three experiments I want to design (三個我想設計的實驗) ::: ## User Journey (從開始用到現在的體驗) Pick a product that you love using (spotify, linkedin, etc) - Describe the user journey of the things you loved about it from the moment you started using it to your use of it now ### First Use I initially signed up for Suno out of curiosity about the power of AI music composition. Because when I work, I listen to some focused music, and I want to try to make my own channel. The process was incredibly fast: enter a text prompt (e.g., "lofi hiphop for studying") → music generated in seconds. This "instant gratification" was my first love. ### Medium-Term Use I then began experimenting with more genres (jazz, blues, swing jazz, ambient) and even create some music with lyrics. I love that I can experiment with different styles like a music producer without a music background. ### Current Use I currently use Suno primarily for: Quickly generating background music (for work/focus/study/relax). Sharing AI-generated music on YouTube, Spotify, Apple Music channel, etc. ## Three experiments I want to design (三個我想設計的實驗) Describe 3 experiments - You should detail out the allocation of each test cell and the different conditions you’re testing - Your hypothesis of which leading and lagging metrics would be impacted by this experiment ### 實驗 1：新手留存 – Custom 教學（Experiment 1: New User Retention – Custom Tutorial） --- #### 假設 (Hypothesis) 新手第一次使用 Suno 的 **Custom 功能**時，常覺得過於複雜，只會用最簡單的 Simple 模式如果提供： - 「+Audio / +Persona / +Inspo」的彈跳真人影片教學 - 「情境選單」彈跳真人影片教學（如健身歌、學習背景音、抖音熱門風格）則能讓新手更快體驗成功，提升留存 When new users first try the **Custom feature* in Suno, they often feel it’s too complex and stick to the simplest Simple mode. If we provide: - Pop-up tutorial videos introducing “+Audio / +Persona / +Inspo” - Contextual pop-up tutorials (e.g., workout songs, study background music, TikTok trending styles) then users will experience success faster, improving retention. --- #### 實驗設置 (Experiment Setup) - 對象：新註冊用戶，首次進入 Custom 功能者 - 平台：Suno official Web - 期間：約 1–2 週，涵蓋至少一個 D7 留存觀測窗 - Target: Newly registered users, first-time Custom users - Platform: Suno official Web - Duration: About 1–2 weeks, covering at least one D7 retention window --- #### 實驗設計 (Test cells) - **Control**：現況（無教學） - **Treatment A**：功能導向影片（介紹 +Audio / +Persona / +Inspo，含範例 prompt） - **Treatment B**：情境導向影片（健身 / 學習 / 短影音情境，含一鍵模板） - **Treatment C**：輕量教練模式（逐步小浮層引導填寫）分配比例：平均分組 (25% each) - Control: Current experience (no tutorial) - Treatment A: Feature-oriented tutorial (introducing +Audio / +Persona / +Inspo, with sample prompts) - Treatment B: Context-oriented tutorial (workout / study / short-video templates, with one-click options) - Treatment C: Light coaching mode (step-by-step tooltips guiding input) Allocation ratio: equal split (25% each) --- #### 指標 (Metrics) ##### Primary (核心目標指標 / Primary Metrics) - 首次成功生成率 (First-Session Success Rate) - 首次成功生成所需時間 (Time to First Success) - Day 1 Custom 使用率 (Day 1 Custom Usage Rate) ##### Leading (前導指標 / Leading Indicators) - 平均生成數 (Day 0–1 Average Generations) - 範例/模板使用率 (Example/Template Usage Rate) - 放棄率 (Abandonment Rate: started filling but didn’t click Generate) ##### Lagging (滯後指標 / Lagging Metrics) - Day 7 留存率 (Day 7 Retention Rate) - Day 7 Custom 重複使用率 (Day 7 Custom Repeat Usage Rate) - 付費轉換率 (Trial → Subscription Conversion Rate) ##### Guardrails (護欄指標 / Guardrail Metrics) - 首屏載入時間 (First-Screen Load Time) - 客訴/錯誤率 (Customer Complaint / Error Rate) ### 實驗 2：手機版版面優化 – 加入 +Audio / +Persona / +Inspo (Experiment 2: Suno Mobile UI Optimization – Adding +Audio / +Persona / +Inspo) --- #### 假設 (Hypothesis) Suno **手機版**目前並沒有「+Audio / +Persona / +Inspo」的入口，導致新手在手機上使用 **Custom 功能**時，多數人僅使用歌詞和風格功能，生成率有限如果在手機版的 **Custom 版面**中： - 加入「+Audio / +Persona / +Inspo」的明顯入口 - 或提供簡化的一鍵選項（如快速添加音色、角色、靈感）則能降低使用摩擦，提升首次生成成功率與整體生成率 Currently, the **Suno mobile app** does not show “+Audio / +Persona / +Inspo” options. This causes new users to mostly stick to the lyric and style functions, limiting generation success. If we update the **Custom UI on mobile** by: - Adding clear entry points for “+Audio / +Persona / +Inspo” - Or providing simplified one-click options (e.g., quick add for audio style, persona, or inspiration) then first-session success and overall generation rate will increase. --- #### 實驗設置 (Experiment Setup) - **對象 (Target):** Suno 手機版 Custom 功能者 - **平台 (Platform):** Suno Mobile (iOS / Android) - **期間 (Duration):** 約 1–2 週，涵蓋至少一個 D7 留存觀測窗 - **Target:** Users on mobile - **Platform:** Suno Mobile (iOS / Android) - **Duration:** About 1–2 weeks, covering at least one D7 retention window --- #### 實驗設計 (Test Cells) - **Control**：現況（無 +Audio / +Persona / +Inspo） - **Treatment A**：在版面中加入「+Audio / +Persona / +Inspo」按鈕（小圖示 + 簡短說明） - **Treatment B**：簡化版入口（如一鍵添加「角色 / 音色 / 靈感」的快捷按鈕） - **Treatment C**：混合設計（顯示「+Audio / +Persona / +Inspo」按鈕 + 精選一鍵模板區）分配比例：平均分組 (25% each) - **Control:** Current layout (no +Audio / +Persona / +Inspo) - **Treatment A:** Add “+Audio / +Persona / +Inspo” buttons in layout (with small icons + short hints) - **Treatment B:** Simplified quick-add options (one-tap Persona / Audio style / Inspiration) - **Treatment C:** Hybrid design (full +Audio / +Persona / +Inspo buttons + curated one-click templates) Allocation ratio: equal split (25% each) --- #### 指標 (Metrics) ##### Primary (核心目標指標 / Primary Metrics) - 首次成功生成率 (First-Session Success Rate) - 首次成功生成所需時間 (Time to First Success) - Day 1 Custom 使用率 (Day 1 Custom Usage Rate) ##### Leading (前導指標 / Leading Indicators) - 平均生成數 (Day 0–1 Average Generations) - +Audio / +Persona / +Inspo 使用率 (Usage Rate of new options) - 放棄率 (Abandonment Rate: started filling but didn’t click Generate) ##### Lagging (滯後指標 / Lagging Metrics) - Day 7 留存率 (Day 7 Retention Rate) - Day 7 Custom 重複使用率 (Day 7 Custom Repeat Usage Rate) ##### Guardrails (護欄指標 / Guardrail Metrics) - 首屏載入時間 (First-Screen Load Time) - 客訴/錯誤率 (Customer Complaint / Error Rate) ### 實驗 3：AI 靈感推薦 – 增加生成深度 (Experiment 5: Personalized Prompts – Encouraging Continued Generation) --- #### 假設 (Hypothesis) 目前部分新用戶只嘗試生成一首歌後便離開，缺乏繼續探索的動機如果在用戶生成完第一首歌後，系統能馬上推薦「下一個可以嘗試的風格 / 人設 / 靈感」例如： - 你剛創作了學習背景音 → 推薦「Lo-fi」或「專注爵士」 - 你剛創作了健身歌 → 推薦「EDM」或「搖滾風格」則能引導用戶持續創作，降低「一次性使用」的情況，提高平均生成數與留存 Some new users generate only one song and then leave, lacking motivation to continue exploring. If, immediately after generating their first song, the system recommends “the next style / persona / inspiration to try” Such as: - After creating study background music → suggest Lo-fi or Focus Jazz - After creating a workout track → suggest EDM or Rock style then users will be encouraged to keep creating, reducing one-time usage and increasing average generations and retention. --- #### 實驗設置 (Experiment Setup) - **對象 (Target):** 新註冊並至少完成 1 次生成的用戶（Web + Mobile） - **平台 (Platform):** Suno official Web & Mobile - **期間 (Duration):** 約 1–2 週，涵蓋至少一個 D7 留存觀測窗 - **Target:** Newly registered users who have generated at least once (Web + Mobile) - **Platform:** Suno official Web & Mobile - **Duration:** About 1–2 weeks, covering at least one D7 retention window --- #### 實驗設計 (Test Cells) - **Control**：現況（生成後無進一步建議） - **Treatment A**：生成後顯示固定建議（3 種常用風格，例如 Lo-fi、Pop、EDM） - **Treatment B**：生成後顯示「依照第一次生成內容」的動態推薦 - **Treatment C**：生成後顯示「動態推薦 + 一鍵生成按鈕」分配比例：平均分組 (25% each) - **Control:** Current experience (no post-generation recommendations) - **Treatment A:** Show fixed suggestions (3 common styles such as Lo-fi, Pop, EDM) - **Treatment B:** Dynamic recommendations based on user’s first generation - **Treatment C:** Dynamic recommendations + one-click generate buttons Allocation ratio: equal split (25% each) --- #### 指標 (Metrics) ##### Primary (核心目標指標 / Primary Metrics) - 平均生成數 (Day 0–1 Average Generations) - 第二次生成率 (Second-Generation Rate) - Day 1 Custom 使用率 (Day 1 Custom Usage Rate) ##### Leading (前導指標 / Leading Indicators) - 推薦點擊率 (Recommendation Click-through Rate) - 一鍵生成使用率 (One-click Generate Usage Rate) - 放棄率 (Abandonment Rate: generated once but left) ##### Lagging (滯後指標 / Lagging Metrics) - Day 7 留存率 (Day 7 Retention Rate) - Day 7 Custom 重複使用率 (Day 7 Custom Repeat Usage Rate) ##### Guardrails (護欄指標 / Guardrail Metrics) - 首屏載入時間 (First-Screen Load Time) - 客訴/錯誤率 (Customer Complaint / Error Rate)