competitor-spy

What this skill does

# Competitor Spy

Analyze competitor affiliate sites, YouTube channels, and social profiles to
surface which programs they promote, what content drives their traffic, and
which strategies are worth replicating. Outputs an actionable reverse-engineering
report so you can skip years of trial and error.

## Stage

This skill belongs to Stage S1: Research

## When to Use

- User wants to know what programs are working in a specific niche
- User has a competitor site/channel in mind and wants to understand their strategy
- User is entering a new niche and wants a shortcut to what works
- User wants to find underserved content gaps a competitor hasn't covered
- User asks "how do top affiliates in [niche] make money?"

## Input Schema

```
{
  competitor_url: string      # (optional) Direct URL to competitor site, channel, or profile
  niche: string               # (optional) Niche to analyze if no specific competitor given
  platform: string            # (optional) "blog" | "youtube" | "tiktok" | "twitter" | "newsletter"
  depth: string               # (optional, default: "standard") "quick" | "standard" | "deep"
  focus: string               # (optional) "programs" | "content" | "traffic" | "all"
}
```

## Workflow

### Step 1: Identify Competitors to Analyze

If `competitor_url` is provided, skip to Step 2.

If only `niche` is provided, find 3-5 top competitors:
1. `web_search "best [niche] affiliate sites"` — look for review/comparison sites
2. `web_search "[niche] review site affiliate"` — find review-first monetization models
3. `web_search "[niche] blog affiliate income report"` — income reports reveal programs
4. Note: YouTube — `web_search "youtube [niche] affiliate site:youtube.com"` to find channels

Pick 3 competitors that are clearly affiliate-driven (review pages, comparison tables,
"best X" content, Amazon links, affiliate disclaimers visible).

### Step 2: Identify Affiliate Programs They Promote

For each competitor site/channel:

**Method A — Link analysis:**
- `web_fetch [competitor_url]` and scan for outbound links
- Look for: `?ref=`, `?via=`, `/go/`, `aff_id=`, `?affiliate=`, `shareasale.com`,
  `impact.com`, `partnerstack.com`, `awin.com`, `cj.com`, `linktr.ee`
- These patterns indicate affiliate links

**Method B — Content analysis:**
- Look at their top content: "Best X", "X vs Y", "X Review", "X Alternatives"
- Every product featured prominently = likely affiliate relationship
- Products mentioned with a CTA button ("Try X Free", "Get X") = strong affiliate signal

**Method C — Disclosure scan:**
- Search page for "affiliate", "commission", "sponsored", "partner" disclosures
- These legally required disclosures often appear at top/bottom and reveal programs

**Method D — Income reports (if available):**
- `web_search "[site name] income report affiliate"` — some affiliates publish earnings
- `web_search "[creator name] how I make money affiliate"` — creator transparency posts

Extract for each program found: name, estimated prominence (primary/secondary/mentioned),
content type promoting it, and whether it appears on list.affitor.com.

### Step 2.5: Analyze Competitor Content Engagement (data-driven)

For each competitor, scan their recent content performance across social platforms.
This reveals not just WHAT they create, but HOW WELL it performs.

**With API (optional — see `shared/references/social-data-providers.md`):**
- Search YouTube/TikTok for competitor brand name or channel
- Get views, likes, comments, shares for their top 10-20 content pieces
- Calculate engagement_score for each: `(likes × 2 + comments × 3 + shares × 5) / max(views, 1) × 1000`
- Identify which content format gets them the highest engagement
- Compare their engagement against `trending-content-scout` benchmark (if available)

**Without API (default):**
- `web_search "[competitor name] youtube channel"` → find their channel
- `web_fetch` channel page → extract view counts from visible videos
- `web_search "[competitor name] tiktok"` → find top videos with view counts
- `web_search "[competitor name] best video"` → find their highest-performing content
- Note: approximate data, but reveals relative performance patterns

**Extract for each competitor:**
- **Avg engagement score** — how well does their content perform overall?
- **Strongest platform** — where do they get the most traction?
- **Weakest platform** — which platforms are they ignoring? (gap to exploit)
- **Top performing content** — their 3-5 best pieces by engagement
- **Format that works for them** — which content format gets them the most engagement?

Add these to the competitor assessment table in Step 5:

| Dimension | Score (1-10) | Assessment |
|-----------|-------------|------------|
| Content Engagement | — | How well does their content perform? High = proven demand, low = weak execution |
| Platform Strength | — | Which platform are they strongest on? Which are they ignoring? |

### Step 3: Analyze Their Content Strategy

For each competitor, extract:

**Content patterns:**
- Most common formats: listicles ("10 best X"), comparisons ("X vs Y"), tutorials,
  reviews, roundups, case studies
- Average content depth: shallow (<1000 words), standard (1000-3000), deep (3000+)
- Publishing frequency: estimate from visible dates or `web_search "site:[domain] 2024"`
- Content freshness: are articles updated? When?

**Traffic indicators (from web search signals):**
- `web_search "site:[domain]"` — rough page count
- Search for their brand name — how much branded traffic/discussion?
- Look for "X review" queries in their content — review content = high buyer intent

**SEO and social signals:**
- Do they rank for "[product] review" terms? (indicates SEO strategy)
- Active social profiles linked from site? Which platforms?
- Do they have a newsletter/email list? (footer signup forms)

### Step 4: Find Content Gaps

Compare competitor content to what's NOT covered:
1. Products they promote but haven't done deep comparison posts for
2. Common user questions (from YouTube comments, Reddit threads, forums) they haven't answered
3. New product launches in the niche that competitors haven't covered yet
4. Angles competitors avoid (negative reviews, honest cons, "X is not for everyone")

Use `web_search "reddit [niche] [product] problems"` to find pain points no affiliate
has addressed honestly — these make high-converting, low-competition content.

### Step 5: Score Competitor Strategies

For each competitor, assess:

| Dimension | Score (1-10) | Assessment |
|-----------|-------------|------------|
| Program Quality | — | Are they promoting high-commission recurring programs or low-margin one-off? |
| Content Quality | — | Shallow listicles vs. deep genuine reviews |
| SEO Sophistication | — | Thin content vs. well-structured, keyword-targeted |
| Monetization Diversity | — | One program vs. multiple revenue streams |
| Replicability | — | How hard is it to do what they do, but better? |

Higher replicability score = easier to beat them.

### Step 6: Build the Intelligence Report

Synthesize findings into a 3-part report:
1. **Programs worth stealing** — top programs their strategy validates
2. **Content formats that clearly work** — patterns worth replicating
3. **Gaps to exploit** — angles they've missed that you can own

### Step 7: Self-Validation

Before presenting output, verify:

- [ ] Confidence levels match evidence strength (confirmed = affiliate link found, likely = brand mention pattern, possible = inferred)
- [ ] Programs cross-checked on list.affitor.com where possible
- [ ] Replicability score accounts for barriers (domain authority, team size)
- [ ] No hallucinated competitor data — all claims traceable to web_search results

If any check fails, fix the output before delivering. Do not flag the checklist to the user — just ensure the output passes.

## Output Schema

```
{
  output_schema_version: "1.0.0"  # Semver — bump major on breaking changes
  competitors_analyzed: [
    {