How to: Make Google Dorks

1. Define Your Keywords

Key points

  • Language – Match the language of your target audience or region. For example, if you are focusing on Poland, your keywords should be in Polish.
  • Country/Region – Adapt your keywords to the country’s context and trends (e.g. Germany vs. United States), where this is relevant to your research goal.
  • Niche/Domain – Choose a niche aligned with your target (e.g. Netflix, Paramount, Doordash...): gaming, streaming, sports, news, etc.
  • Source Freshness – Prefer sources that publish recent content (e.g. new games on Steam, current news, or trending topics on social media).

Before generating keywords, make sure to clarify what's the desired outcome. For instance, if your goal is to get outlook lines you won't make keywords like "outlook login", "hotmail sign in". Instead:

  • Identify which countries or regions are most relevant to your target (outlook would be worldwide but mostly US/CA/EU/JP).
  • Estimate your audience, broad consumer base or a narrow industry vertical like sports for example.
  • Align niches accordingly (e.g. gaming, streaming, news, sports).

To expand your base keyword list, you can use:

  • Google Trends to identify current topics and queries users are searching for.
  • Industry platforms (e.g. app stores, game libraries like steam, streaming catalogs) to collect titles and terms.
  • News and media sites for fresh, topical keywords.

Within Google Trends, you can:

  1. Set the country or region.
  2. Select an appropriate time range (a week or month is often sufficient).
  3. Collect related queries and topics that match your niche.

Aim to build a keyword list of at least 50–100 terms for meaningful coverage. You can then use our keyword scraper to collect related terms around your base keywords. Pull up the DorkPlus dashboard and start a new Keyword Scraper task, for the settings: 1 min. word, 3-4 max. words, 2-3 loops. Then save your filtered keywords.

2. Working With Parameters

Now that you got a bunch of keywords, make a new Scraper task, for the list select your freshly scraped keywords, let the task finish and save the filtered URLs as a file (locally). We'll then use the URLs utility from the Utilities page, upload your links and hit Extract parameters. Then save those parameters.

Clean any extracted parameter lists by removing unrealistic, overly long, or clearly irrelevant entries.

3. Designing Dork Types (Search Query Patterns)

If you have no clue what a search operator is you definitely have to learn about it, the following guides will give you an idea of what those are:

Design “dork types” as reusable search patterns that combine:

  • Relevant keywords.
  • Site-specific filters (e.g. site:example.com).
  • Basic operators (e.g. intitle:, "exact phrases").

Keep these dork types simple, no need to overcomplicate things. The AI dork types builder is useful to create a base but you'll need to manually test/edit these to actually keep the good ones. The process is usually the following:

  • Create 2-5 dork types
  • Generate a small dork list (5-10K lines) with your keywords, parameters and new dork types
  • Run that list on the scraper and monitor the links/dorks ratio, if it's over 2 it's fine, over 4 it's good, over 8 it's perfect, under 2 though you can get rid of those
  • For the page types always use PHP

Start with small lists of dork types to evaluate quality:

  • Group dorks into sets of a few types (e.g. 3–5 template patterns, the lowest the easiest it'll be to find the good dork types).
  • Run small dork list, usually 5-10K dorks will be enough
  • Evaluate:
    • Links/dorks ratio (aim for at leat 2 but preferably 4)
    • Relevance of results
    • Freshness of content

5. Iterating on Poorly Performing Dorks

If a batch of dorks underperforms:

  • Analyze which patterns are producing a low ratio (links/dorks) and get rid of those.
  • Create new dork types with better-focused operators (intitle, site, ext...).
  • Re-run tests with refined dork types.

Treat this as an optimization loop rather than a one-time task. A couple tests will ensure you isolate the proper dork types.

6. Scaling Up High-Performing Dorks

When you identify dork types that consistently yield relevant, high-quality results:

  • Keep these patterns as your core dork types.
  • Gradually increase the size of your dork lists.
  • Periodically re-evaluate performance, especially as trends and content change over time.

When you found a good batch of dork types, at least 5, you can generate a dork list that's about 250-500K lines and run it. Once you scraped links it's recommended to switch to new keywords (new niche, new language perhaps). The parameters may need to be changed if the ratio (lines/dorks) is going down.

7. Tips & Advices

  • Don't generate lists over 1M lines, it's pointless, after 1M lines you'll scraped 50-70% of results you already scraped, they'll get removed by the local antipublic system but it'll just waste time
  • If you're completely new to this dorking world I'd strongly recommend you learning about the operators, Guide here
  • There is no speed running, it'll take an hour or maybe two to get the idea, and a few more hours to get the right dork types
  • AI is good but it has its limits, of course you could just generate keywords/dorktypes/parameters with ChatGPT but you'll see poor results, reason for that is ChatGPT or any AI can only produce what it's trained on, meaning you'll get generic results
  • Manual work is required, you'll have to clean the dork types and parameters manually, learn how the operators work, pick your niche, select some base keywords, but this ensure you'll get the best results

Previous Page

Getting Set Up

Next Page

Beginner Guide

On this page