Failed Projects & Post-Mortems

A legal document and policy generator on onestopwrite.com (privacy policies, terms, cookie policies, business letters) that originated from a domain flip attempt, pivoted to pay-per-use, and was eventually retired.

I originally purchased onestopwrite.com intending to flip the domain. When that didn't work out, I built a web-based policy generator, hypothesizing that providing instant legal policies for website builders and startups would attract search traffic and pay-per-use revenue.

Expected pay-per-use document purchases or substantial organic search volume by providing customizable policy generators.

Pay-per-use pricing attracted zero buyers. Transitioning the tool to 100% free brought 3–8 daily organic users who actively generated policies. However, despite running multiple SEO, AEO (Answer Engine Optimization), and GEO (Generative Engine Optimization) experiments, traffic remained stuck in single digits.

The legal document niche is overwhelmingly dominated by high-authority incumbents with massive backlink moats (e.g. TermsFeed, Termly). The single-digit daily traffic volume did not justify continued maintenance or domain renewal, so I allowed the domain to expire.

A blog website on onestopread.com that was pivoted into an AI-automated government exam alert portal with Gemini Flash summarization and cron-driven scrapers.

Initially thought publishing articles on trending topics would attract organic search traffic. After that failed to rank against high-DA media sites, I pivoted to government exam notifications—hypothesizing that automated scrapers and Gemini Flash could generate clean, structured exam updates faster than cluttered legacy portal competitors.

Expected automated programmatic exam coverage to rank quickly for low-competition notification keywords and build compounding search impressions.

Phase 1 (trending blog) got 0–1 clicks/day. Phase 2 (AI exam portal) showed promising early traction, growing to 20–30 daily clicks across Google and Bing. However, after 20–25 days, clicks abruptly collapsed to 0 and daily impressions flatlined to 5–20. Manual content rewrites with verified official government links failed to recover search rankings.

The domain suffered an algorithmic content penalty or sandbox de-valuation typical of fresh domains running programmatic AI pipelines. Recovering search visibility without substantial external backlink authority was unsustainable, so I abandoned the project.

An AI-powered resume builder, JD optimizer, and interview prep platform with 50–80+ templates hosted on zyoraai.com.

Hypothesized that combining job description (JD) keyword tailoring, AI cover letter generation, and interview question preparation into 50–80+ curated resume templates would convert job applicants into paying subscribers.

Expected organic job-search queries and career communities to drive consistent freemium signups and paid template exports.

Built the complete platform (JD parser, cover letter writer, interview prep, 50–80+ templates including 8–10 tailored for large tech companies) by combining LLM tooling (ChatGPT, DeepSeek, Claude, Antigravity) with Next.js. Deployed on zyoraai.com. However, across 2–4 months of experimentation, organic user acquisition stalled and paid conversion was near zero in a crowded category.

The AI resume builder niche became saturated with heavily-funded incumbents. Organic keywords required massive backlink authority that a new domain could not achieve quickly without unsustainable ad spend. Dropped the product and repurposed the domain for a computer vision project.

An ad-free Android mobile app providing curated semester exam preparation materials, previous year question papers, and educational notes for college students.

College students consistently struggle to find organized, reliable semester syllabus notes and previous year question papers. Most existing resources were scattered across disorganized drives or buried behind ad-heavy websites.

Expected steady organic adoption across university cohorts by offering a clean, completely free mobile app with zero ads.

The app achieved organic traction, reaching 1,000+ downloads on the Google Play Store and actively helping students prepare for semester exams.

In 2024, during my final college year, I joined Mobiloitte in Pune as a Software Developer (left after 3 months due to lack of active project allocation) and then joined Collegedunia as a Software Developer. Due to full-time engineering commitments, I was unable to dedicate time to push required Android target SDK and Play Store policy compliance updates. After 3 warning notices from Google Play, the app and developer account were suspended.