Surround Sound SEO Strategy：我们如何用真实数据构建 SERP 霸占系统

Visibility Score % = (mentions in top 20 ÷ 20) × 100

Weighted Score = Σ(Position Weight where mentioned) / Σ(All Position Weights) × 100

Position Weights: #1 = 10pts, #2 = 9pts, ... #10 = 1pt

Opportunity = (100% - Visibility Score) × Monthly Search Volume

┌─────────────────┐     ┌──────────────────┐     ┌───────────────────┐
│  Keyword         │────▶│  SERP Analyzer    │────▶│  Mention Checker  │
│  Research        │     │  (Top 20 per KW)  │     │  (Scrape + Match) │
│  keyword_        │     │  serp_            │     │  mention_         │
│  research.py     │     │  analyzer.py      │     │  checker.py       │
└─────────────────┘     └──────────────────┘     └───────────────────┘
                                                          │
                                                          ▼
┌─────────────────┐     ┌──────────────────┐     ┌───────────────────┐
│  Outreach        │◀────│  Opportunity      │◀────│  Visibility       │
│  Pipeline        │     │  Scoring          │     │  Calculator       │
│  (7,322 targets) │     │  (Gap Analysis)   │     │  visibility_      │
│                  │     │                   │     │  calculator.py    │
└─────────────────┘     └──────────────────┘     └───────────────────┘

CREATE TABLE keywords (
    id SERIAL PRIMARY KEY,
    keyword VARCHAR(500) NOT NULL,
    silo VARCHAR(100),           -- Content silo grouping
    volume_ahrefs INTEGER,       -- Monthly search volume
    difficulty INTEGER,          -- Keyword difficulty (0-100)
    cpc NUMERIC(10,2),          -- Cost per click
    search_intent VARCHAR(50),   -- informational, commercial, transactional
    country VARCHAR(5),          -- us, de, fr, etc.
    language VARCHAR(5),         -- en, de, fr, etc.
    UNIQUE(keyword, country, language)
);

CREATE TABLE visibility_scores (
    id SERIAL PRIMARY KEY,
    keyword_id INTEGER REFERENCES keywords(id),
    brand_name VARCHAR(100),
    visibility_score NUMERIC(5,2),  -- 0-100%
    mention_count INTEGER,
    pages_with_mentions INTEGER,
    total_pages_checked INTEGER,
    calculated_at TIMESTAMP DEFAULT NOW()
);

# Simplified mention detection logic
def find_brand_mentions(content, brand_names):
    mentions = []
    for brand in brand_names:
        pattern = re.compile(
            rf'\b{re.escape(brand)}\b',
            re.IGNORECASE
        )
        for match in pattern.finditer(content):
            # Extract surrounding context (200 chars)
            start = max(0, match.start() - 100)
            end = min(len(content), match.end() + 100)
            context = content[start:end]
            mentions.append({
                'brand': brand,
                'position': match.start(),
                'context': context,
                'has_link': check_for_link(content, match)
            })
    return mentions

keywords-explorer-overview    → Keyword metrics (volume, difficulty, CPC)
keywords-explorer-matching    → Related keyword discovery
serp-overview                 → SERP position data for any keyword
site-explorer-organic-keywords → Competitor keyword analysis
batch-analysis                → Bulk domain metrics

search_engine       → Live SERP scraping (Google, Bing, Yandex)
search_engine_batch → Batch SERP scraping
scrape_as_markdown  → Full page content extraction
scrape_batch        → Batch content scraping