Google Autocomplete Keyword Research Tool

A multi-depth scraper that harvests Google Autocomplete suggestions from any seed keyword. Expand a single topic into hundreds of long-tail keyword ideas by traversing the full alphabet and recursively following the most relevant suggestions.

---

What the Actor Does

When you type a query into Google, the search bar offers autocomplete suggestions that reflect real user intent. These suggestions are a goldmine for keyword research, content planning, and SEO analysis — but collecting them at scale requires automation.

This Actor automates that process. You provide one or more seed keywords, and the scraper systematically queries Google Autocomplete by appending every letter of the alphabet (a-z) and every digit (0-9). It then takes the suggestions returned at each step and uses them as new starting points, repeating the alphabet sweep until it reaches the configured depth limit. The result is a tree of keyword variants, each tagged with its depth level, the original seed keyword it descended from, and its relevance score from Google.

---

Use Cases

• SEO & Content Strategy: Feed a product or topic into the Actor and receive a structured list of long-tail keywords with relevance scores. Use the output to identify content gaps, plan blog posts, or discover underserved search queries.
• PPC & Ad Campaign Research: Discover the exact phrases users type when searching for a niche. The depth expansion reveals question-style queries and modifier variations that manual brainstorming would miss.
• Market Intelligence: Seed the Actor with competitor brand names or industry terms to map the competitive landscape. The suggestion_type field helps distinguish informational queries from navigational or transactional ones.
• Product & Feature Discovery: Use autocomplete data to surface features, use cases, or integrations that customers mention — e.g., searching "Slack" reveals suggestions like "Slack integration with Jira" or "Slack alternative for teams."

---

How the Search Works

The Actor operates in two distinct phases controlled by the Max Depth parameter:

Phase 1 — Alphabet Expansion (Depth 1)

Every seed keyword is expanded by appending each character in the alphabet and digits:

seed: "python"
depth 0: "python"
  → "python a"  → yields: ["python api", "python course a1", ...]
  → "python b"  → yields: ["python basics", "python blog", ...]
  → "python c"  → yields: ["python conda", "python code", ...]
  → ...
  → "python z"  → yields: ["python zip", ...]
  → "python 0"  → yields: ["python 099", ...]
  → ...
  → "python 9"  → yields: ["python 9x10", ...]

Phase 2 — Recursive Suggestion Expansion (Depth 2)

The suggestions collected in Phase 1 become new seed keywords. Each one is then expanded with the full alphabet again:

suggestion from Phase 1: "python api"
depth 1: "python api a"  → yields: ["python api async", ...]
depth 1: "python api b"  → yields: ["python api basics", ...]
depth 1: "python api c"  → yields: ["python api cookies", ...]
...

Visual Flow

┌─────────────────────────────────────────────────────────┐
│  USER INPUT                                             │
│  keyword = ["python"]                                   │
│  maxDepth = 2                                           │
└──────────────────────────┬──────────────────────────────┘
                           │
                           ▼
            ┌─────────────────────────────┐
            │  PHASE 1: Alphabet Sweep    │
            │  (depth = 0 → 1)            │
            │                             │
            │  "python" + {a..z, 0..9}    │
            │       ↓                     │
            │  36 queries → ~200+         │
            │  suggestions collected      │
            └────────────────┬────────────┘
                             │
                Suggestions become seeds
                             │
                             ▼
            ┌─────────────────────────────┐
            │  PHASE 2: Recursive Sweep   │
            │  (depth = 1 → 2)            │
            │                             │
            │  "python api" + {a..z, 0..9}│
            │  "python basics" + {a..z..} │
            │  "python blog" + {a..z..}   │
            │       ↓                     │
            │  thousands of queries →     │
            │  10,000+ suggestions        │
            └────────────────┬────────────┘
                             │
                             ▼
            ┌─────────────────────────────┐
            │  OUTPUT: Flat Dataset       │
            │                             │
            │  { seedKeyword, query,      │
            │    depth, suggestion,       │
            │    relevance,               │
            │     suggestionType}         │
            │                             │
            └─────────────────────────────┘

Each record in the output carries three key fields:

Field	Description
seedKeyword	The original keyword you provided. Links every result back to your starting point.
query	The exact query sent to Google (e.g., "python api async").
depth	The expansion level. 1 = direct child of your seed. 2 = grandchild. Higher values mean more granular long-tail variants.

---

Input

Parameter	Type	Description
Seed Keywords	Array of strings	One or more starting keywords. The Actor will expand each one through the full search tree.
Max Depth	Integer (1 or 2)	Controls how many levels the Actor follows. Depth 1 collects only direct alphabet expansions. Depth 2 recursively expands the suggestions from Depth 1.
Max Suggestions Limit	Integer	Hard cap on the total number of suggestions collected. The Actor stops once this limit is reached. Defaults to 100.
Target Country	String	ISO 3166-1 alpha-2 country code to localize autocomplete results. Defaults to DE (Germany). Changing this affects which country-specific suggestions Google returns.

---

Output

Each record represents one autocomplete suggestion discovered during the sweep:

{
  "seedKeyword": "python",
  "query": "python api async",
  "depth": 2,
  "suggestion": "python api async",
  "relevance": 680,
  "suggestionType": "QUERY"
}

---

Pro Tips

The Actor can generate thousands of suggestions per seed keyword. To use it efficiently and avoid wasting runs on irrelevant results, follow this workflow:

1. Start with maxDepth = 1

Run the Actor with a shallow depth first. This collects only the direct alphabet expansions of your seed keyword — typically 100–200 suggestions. These are the most broadly relevant results.

2. Review and select

Examine the suggestions from the Depth 1 run. Identify which ones are worth exploring further — for example, "python api" or "python course" if you're building a tutorial site.

3. Seed those back in

Take the promising suggestions from Step 2 and paste them as the new Seed Keywords in a second run. This time set maxDepth = 2 to expand them into long-tail variants.

4. Iterate

Repeat the cycle as needed. Each run narrows the focus and uncovers more specific keyword variants. This iterative approach prevents you from collecting thousands of irrelevant suggestions and keeps compute costs low.

---

Notes

• Proxy usage is automatic. The Actor uses Apify's built-in residential proxy with IP rotation by default, so you don't need to configure anything manually.
• No login required. The Actor does not need Google account credentials — it uses the public autocomplete endpoint.
• Results vary by region. Google Autocomplete is personalized based on location and search history. Use the Target Country parameter to specify which country's autocomplete suggestions you want to collect.

---