{
  "id": 19499696,
  "title": "How One Owner Built a Google Ads Campaign With Claude Code for $0.46",
  "url": "https://urgent.news/2026/11/11/how-one-owner-built-a-google-ads-campaign-with-claude-code-for-0-46",
  "topic": "ai",
  "section": "AI",
  "published": "2026-11-11T13:17:40.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/fun_fun_19313eced54af3c68/how-one-owner-built-a-google-ads-campaign-with-claude-code-for-046-kg7"
  },
  "original_language": "en",
  "account": "A Texas car detailing business owner enlisted Claude Code to research, write, and create a Google Ads search campaign for $0.46. Here's how they did it, with the costs of each step and key takeaways. The aim is not that ads are now free, but that the research layer that previously required expensive subscriptions can now cost under a dollar per campaign. The problem, however, is that the ads did not bring in any clients. The owner had previously created ads manually, researching keywords and setting up campaigns. Those campaigns didn't yield clients, so they tried Semrush's cheapest plan, found it too limited, and abandoned it. First attempt: Claude Code without data. With just a Claude Code Pro subscription and no ad tools, the owner generated a generic plan based on their existing knowledge. Mobile detailing, interior car detailing, and truck detailing were among the services. However, the plan lacked verified search volumes, costs, and competitor data. Second attempt: Claude Code with an MCP and real data. The owner connected an MCP server (using the AIML API MCP) promising an optimized campaign for under $1. The prompt was simple: connect the MCP and create an optimized ad campaign for the company, with the website link provided. Claude Code then listed the steps it took and their costs: checking the landing page, retrieving keyword ideas from the site and 20 seed terms, finding rival advertisers, pulling their ads, checking which ads show on live searches, sorting keywords into buyers, other brands, and junk, forecasting clicks and cost, writing ad groups, ads, and negative keywords. The agent then set up the whole campaign in Google Ads and asked for the owner's approval before launching. The keyword plan produced by Claude Code included a tiered list of keywords with Texas search volume and CPC. Tier 1 (core terms) focused on the most budget-friendly keywords. One caveat was that many people searching for \"car detailing near me\" were actually looking for a shop, not a mobile service. The agent recommended adding \"mobile\" and \"we come to you\" to the ad copy and closely monitoring the search terms report. Tier 2 (cheaper and useful) keywords included truck detailing, engine bay cleaning, car wash and detail, and price-shopper terms like car detailing prices and mobile detailing cost. These terms were only worth pursuing if the website showed prices. Tier 3 (test small) keywords included paint correction and ceramic coating, but these should only be tested if the service is offered. Community members pointed out the need for local context, noting that data could miss specific searches in towns with smaller populations. The owner emphasized that the $0.46 covered only the research and setup costs, not the actual ad spend. After creating the campaign, the owner tracked clients, not just clicks, by monitoring calls, bookings, and the search terms report. Adjustments were made based on these metrics.",
  "summary": "A local car detailing business owner in central Texas had Claude Code research, write, and set up an entire Google Ads search campaign, and the data costs came to $0.46 . Here is what the agent actually did, what each step cost, and the guardrails worth copying. (h/t the owner who shared the original write-up.) The point is not that ads are now free. The point is that the research layer that used…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}