amazon-ppc-campaign

# Amazon PPC Campaign Optimization 📢

Build profitable PPC campaign structures from scratch, or audit and optimize existing campaigns with data-driven bid adjustments. No API key — works out of the box.

## Installation

```bash
npx skills add nexscope-ai/Amazon-Skills --skill amazon-ppc -g
```

## Two Modes

| Mode | When to Use | Input | Output |
|------|-------------|-------|--------|
| **A — Build** | Launching PPC for a new product | Product info + keywords + margins | Complete campaign blueprint + keyword groupings + initial bids |
| **B — Optimize** | Improving existing campaigns | Campaign data + search term reports + current ACoS | Optimization plan + bid adjustments + negative keyword list |

## Capabilities

- **ACoS financial framework**: Calculate break-even ACoS, target ACoS, and Max CPC from product margins — the foundation for every bid decision
- **Campaign architecture design**: Build a structured Auto → Broad → Exact funnel with proper negative keyword isolation between campaigns
- **Keyword grouping**: Organize keywords into campaign buckets with match types and initial bids based on confidence level
- **Bid optimization**: Apply ACoS-based bid adjustment rules using industry-standard formulas (cut/increase by percentage based on ACoS range)
- **Keyword funnel analysis**: Identify migration opportunities (Auto→Broad→Exact) and wasted spend (high-click zero-sale terms)
- **Negative keyword management**: Generate seed lists (cross-campaign, irrelevant terms, generic waste modifiers) and ongoing additions from search term data
- **Search term report analysis**: Parse user-provided campaign data to find profitable terms, wasteful terms, and optimization gaps
- **Competitor ASIN targeting**: Build product targeting campaigns aimed at competitor product pages
- **Integration chain**: Works with [amazon-keyword-research](https://github.com/nexscope-ai/Amazon-Skills/tree/main/amazon-keyword-research) for keyword input and [amazon-listing-optimization](https://github.com/nexscope-ai/Amazon-Skills/tree/main/amazon-listing-optimization) for pre-launch listing quality checks

## Usage Examples

### Mode A — Build New Campaigns

```
I'm launching a portable blender on Amazon US. Price: $39.99. Product cost: $8, shipping: $3, Amazon fees: $7.50. Here are my keywords: portable blender, personal blender, smoothie maker. Build me a PPC campaign structure.
```

```
Use amazon-keyword-research to find keywords for "bamboo cutting board", then build a PPC campaign structure. Product costs $6, sells for $29.99. Brand new product launch.
```

```
I want to advertise my dog t-shirt (ASIN B0D72TSM62, price $5.99, cost $2). Look at competitors B0CMD17929 and B0B76519ZG, extract their keywords, and build my PPC campaigns.
```

### Mode B — Optimize Existing Campaigns

```
My PPC ACoS is 58% and my target is 30%. I have 3 campaigns: Auto ($800/month, ACoS 67%), Manual Broad ($1,100, ACoS 48%), Manual Exact ($500, ACoS 33%). Product margin is 54%. Help me optimize.
```

```
Here's my search term report [paste CSV data]. Break-even ACoS is 40%. Find wasted spend, tell me what to negate and what to migrate.
```

```
Weekly PPC check: here are this week's search terms with clicks and sales [data]. Add negatives for 10+ clicks with no sales, move 2+ orders to Exact.
```

### Short Prompts Work Too

```
Help me set up PPC for my product B0D72TSM62
```
```
My ACoS is too high, help me fix it
```
```
I want to start advertising on Amazon
```

---

## How This Skill Collects Information

Users rarely provide everything upfront — and they don't need to. This skill follows a **progressive information gathering** approach:

**Step 1: Extract from the prompt.** Parse whatever the user already provided — ASIN, price, ACoS numbers, campaign names, keywords, etc.

**Step 2: Auto-discover.** If an ASIN is given, run the bundled `scripts/fetch-competitor.sh <ASIN>` to get price, category, BSR, and competitor context. This script handles Amazon's anti-bot protections. If the user mentions a product type without an ASIN, use `web_search` to understand the market.

**Step 3: Identify gaps.** Compare what you have against what's needed (see the Required Information tables in Mode A Step A1 and Mode B Step B1 below). Focus on what's **critical** to proceed:
- Mode A critical: **product costs** (to calculate ACoS) + **monthly ad budget** (to size campaigns) + **keywords or competitor ASINs** (to build campaigns)
- Mode B critical: **current ACoS** + **profit margin** (to know the gap and set targets)

**Step 4: One consolidated follow-up.** Ask only for missing critical items — in one conversational message, not a questionnaire:

```
Mode A example:
"I found your product — Paiaite Dog T-Shirt, $5.99 on Amazon. To build your
campaigns, I need three things:
  1. Your product cost per unit (so I can calculate your break-even ACoS)
  2. Your monthly ad budget (so I can size the campaigns right)
  3. Any target keywords or competitor ASINs? (Or I can research for you)"

Mode B example:
"Got it — ACoS is too high. To give you specific actions, can you share:
  1. Your profit margin (or product cost, I'll calculate it)
  2. Which campaigns are running and their rough ACoS?
Search term report data is a bonus but not required to start."
```

**Step 5: Use estimates when stuck.** If the user can't provide something (e.g., doesn't know exact fees), use reasonable category-based estimates and clearly note the assumption. Never block progress waiting for perfect data.

---

## Key Concepts

Three formulas drive every recommendation in this skill. They're introduced here and applied in Step A2 (for Mode A) and Step B2 (for Mode B).

**Break-even ACoS** = Profit margin before ad spend. If your product sells for $40 with $15 in costs after Amazon fees, your margin is $25/$40 = 62.5%. At 62.5% ACoS you spend all profit on ads — break even.

**Target ACoS** = Break-even ACoS − Desired profit margin. Want 25% profit after ads? Target