# Douyin Engagement Algorithm AI Agent Connect

> Calculate precise video engagement scores and viral potential for Douyin and TikTok content.

## Overview
- **Category:** marketing
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_yvHJybH5kNWsnuAjyRTBthxCOZewqtWBgUM2uEoZ/ai-agent-connect
- **Tags:** douyin, tiktok, engagement, analytics, viral

## Description

This MCP server provides deterministic tools to quantify video performance in the Douyin and TikTok ecosystems. It calculates a weighted engagement score normalized per 1,000 views, allowing for fair comparison across different video scales. Use `calculate_engagement_score` to determine raw quality, `predict_viral_potential` to find the growth tier, `calculate_posting_adjustment` to account for freshness decay, and `estimate_view_multiplier` to forecast future reach.

## Tools

### calculate_engagement_score
Determine the raw engagement quality of a video relative to its reach

### calculate_posting_adjustment
Adjust the expected engagement based on the time elapsed since the video was posted

### estimate_view_multiplier
Forecast the potential reach of a video based on its current performance

### predict_viral_potential
Categorize a video's growth trajectory into a standardized index

## Prompt Examples

**Prompt:** 
```
Calculate the engagement score for a video with 5000 views, 80% completion rate, 200 likes, 50 comments, 30 shares, and 10 saves.
```

**Response:** 
```
The calculated engagement score is 34.5 per 1,000 views.
```

**Prompt:** 
```
What is the viral potential for a video with an engagement score of 85?
```

**Response:** 
```
The video has a Viral Potential Index of 92 and is in the 'Explosive' growth tier.
```

**Prompt:** 
```
Estimate the view multiplier for a video with 10,000 views and an engagement score of 40.
```

**Response:** 
```
The predicted view multiplier is 5.2, suggesting significant growth potential.
```

## Frequently Asked Questions

**How is the engagement score calculated?**
The score is calculated using a weighted formula: (completion_rate × 0.4) + (shares × 0.3) + (saves × 0.15) + (comments × 0.1) + (likes × 0.05), all normalized per 1,000 views.

**Can I predict if a video will go viral?**
Yes, by using the `predict_viral_potential` tool, you can obtain a Viral Potential Index from 0 to 100 and a specific growth tier.

**Does the algorithm account for video age?**
Yes, the `calculate_posting_adjustment` tool applies a historical decay rate to adjust engagement scores based on the hours elapsed since upload.