Case Study: SEO aIdeas – Autonomous Marketing Auditor

Case Study: SEO aIdeas – How technology wins over empty reports

Most SEO agencies send clients automatically generated reports from commercial tools (e.g. Senuto/Semstorm) for which the client pays thousands of PLN. The SEO aIdeas project came out of frustration with that business model. I wanted to see whether a simple, database-less Python script combined with AI could deliver deeper analytics to a company's board in seconds than an agency's analyst team could in a week.

Client deployment outcome (e-commerce, audioplaza.pl): The time needed to analyze positions for 53 key brands and detect SEO weak spots fell from 16 hours of manual work to just 12 seconds. The system detected critical brand gaps on 248 keywords.

1. The Business Problem

The client has over 280 brands and 20,000 products in stock. Monitoring organic visibility for such a large scale of assortments in traditional tools was either extremely expensive (keyword limits) or required hours of manually copy-pasting data into Excel.

Marketing agencies reported an overall increase in site visibility, hiding the fact that on the most important buying-intent keywords (e.g. "amplifier + brand"), the store was losing rankings to smaller competitors, and organic traffic was growing mainly on the store's own brand terms.

2. System Architecture (Aegis Agentic Pipeline)

I designed and implemented a fully serverless and database-less data pipeline written in Python 3:

Ingestion Loop: The script extracts brand and product model names from the store's sitemap, then periodically queries the Serper API (a stable Google search scraper) for search engine positions.
Brand Gap Detection: A heuristic engine processes CSV files, identifying core brand phrases that are outside the TOP 10 (yielding no traffic), despite products being available in the store.
Competitor Analysis (Nemesis): The algorithm identifies domains that most frequently push our store off the first page of search results and points out their strengths.
AI Recommendation Generation (Claude API): Compiled gap and opportunity data (Quick Wins from positions 11-20) are sent to the Claude 3.5 Sonnet model to automatically generate personalized optimization instructions for developers and copywriters.

3. Sample Code Snippet

The simplified code below shows how Aegis fetches rankings from the Serper API in just a single request per keyword, minimizing API credit consumption:

import requests
import os

def check_google_ranking(query, target_domain):
    # Querying Serper API for google.pl results
    api_key = os.environ.get("SERPER_API_KEY")
    headers = {"X-API-KEY": api_key, "Content-Type": "application/json"}
    payload = {"q": query, "gl": "pl", "num": 30}

    response = requests.post("https://google.serper.dev/search", json=payload, headers=headers)
    results = response.json().get("organic", [])

    for item in results:
        link = item.get("link", "")
        if target_domain in link:
            return item.get("position") # Returns position (e.g. 4)
    return None # Outside TOP 30

4. Business Conclusions

By replacing expensive SQL databases with flat CSV and JSON file structures, the system incurs zero server maintenance costs. The client gained independence from the agency – at every weekly meeting, the board has an objective report showing which actions yielded results and which were merely faked.

Want to implement similar analytical controls in your business?

I will build a custom technical audit system for your company without monthly licensing fees.

Consult system implementation