Comment Analysis

TubeLog's Comment Analysis (available on Pro plans and above) uses AI to read and synthesize your video's comment section, turning thousands of individual comments into actionable insights.

#What Comment Analysis Does

For any video with more than 10 comments, TubeLog generates:

1. AI Summary — a concise 3–5 sentence paragraph capturing the overall tone and key themes from the comments
2. Sentiment Breakdown — percentage split between positive, neutral, and negative sentiment
3. Key Themes — recurring topics, questions, and feedback clusters
4. Improvement Suggestions — specific, actionable improvements derived from critical comments

#Reading the AI Summary

The summary appears at the top of the Comments tab:

"Viewers responded enthusiastically to the tutorial structure, especially the step-by-step breakdown. Multiple comments requested a follow-up video on advanced techniques. The main criticism was that the audio levels dropped noticeably around the 4-minute mark. Several viewers shared their results in the replies, indicating high engagement and implementation."

This summary is regenerated whenever you refresh comment analysis (counts toward daily API limits — see below).

#Sentiment Breakdown

| Category | What It Captures | |----------|-----------------| | Positive | Praise, gratitude, enthusiasm, shares | | Neutral | Questions, observations, requests | | Negative | Criticism, complaints, confusion |

A healthy video typically has 70–85% positive, 10–20% neutral, and under 10% negative. Higher-than-normal negative rates often indicate:

• Misleading thumbnail or title
• Audio/video quality issues
• Content didn't deliver on its promise
• Controversial topic

#Filtering Comments

Use the filter bar to drill into specific subsets:

#By Sentiment

• Positive only — find shareable praise and testimonials
• Negative only — prioritize which issues to address
• Questions only — find viewer questions you haven't answered yet

#By Reply Status

• Unreplied — quickly find comments waiting for a response. Consistent replies increase engagement score and build community loyalty.

#By Keyword

Type any word to find all comments mentioning it. Useful for:

• Finding mentions of specific timestamps ("the part at 3:30")
• Checking if viewers noticed a specific element
• Moderating for spam or policy violations

#Random Comment Picker

Running a giveaway or selecting a winner from comments? The Comment Picker:

1. Filters eligible comments (e.g., must include a keyword like "entered")
2. Optionally requires account age > X days (anti-spam)
3. Selects a winner randomly with full audit trail
4. Shows runner-up suggestions

Access via Comments → ⋮ → Pick a winner.

#API Usage and Limits

Comment analysis uses the YouTube Data API v3. There are per-video limits:

| Plan | Comments fetched per refresh | Refreshes per day | |------|------------------------------|------------------| | Pro | 500 most recent | 5 | | Enterprise | 2,000 most recent | 20 |

Note: For videos with more than 500 comments, the AI summary is based on a representative sample. TubeLog prioritizes recent comments and highly-liked comments in the sample.

#Comment Moderation Actions

From the Comments tab, you can:

• Reply — post a reply directly from TubeLog (requires write permission)
• Like comment — like a comment to show appreciation
• Hide user — hide all comments from a specific user
• Report — flag a comment for YouTube review

These actions require the youtube.force-ssl write permission scope. If you only granted read permissions during channel setup, you'll be prompted to re-authorize with additional scopes.