iSearchFrom

Our initial goal was to validate whether users of our free legacy tool (a simple, proxy-less URL modifier) would pay for a more advanced, premium proxy search tool.

Partnering closely with one of the core stakeholders, we took a scrappy approach to research. We spent hours diving into specialized SEO subreddits and followed up with direct interviews with professional marketers to uncover their frustrations with existing solutions.

While we successfully validated the demand for a robust proxy search, our synthesis revealed a massive, unexpected shift in the broader search landscape. Marketers were seeing significant traffic migrating to AI chatbots and needed a way to track it just like they track Google results.

This insight completely expanded our product strategy. We realized we couldn't just build an advanced proxy tool for Google; we also needed to build parallel AI tracking tools capable of monitoring organic references and anticipating the inevitable rollout of chatbot ads.

The research proved there was a strong appetite for a paid SaaS platform—one that combined the reliability of an enterprise proxy search with cutting-edge AI visibility tracking.

The core challenge of the SaaS application was data density. The old interface allowed users to configure searches instantly. The new dashboard needed to maintain frictionless execution while introducing highly complex tracking mechanics and AI visibility reports.

I focused heavily on the SaaS workspace—moving away from a simple form to a comprehensive dark-mode optimized application where users could save configurations, collaborate with teams, and monitor historical search data.

Every interaction was scrutinized to ensure it felt like an enterprise-grade tool. I prioritized keyboard-first navigation and zero-latency execution, knowing that power users would immediately reject a UI that slowed them down.

iSearchFrom users need systems that respect their time and expertise. I focused on high-density ergonomics at the micro-level:

• Keyboard-First Navigation: Allowing veterans to navigate search configurations and toggle views instantly without touching the mouse.

• Data Density Control: Users can toggle between comfortable and high-density views for data tables, ensuring the UI adapts to their specific workflow.

• Frictionless Execution: The core search simulator was refined to remove any latency, ensuring searches execute instantly with exact geolocation parameters.

We initially launched an MVP focused on the new premium Search Engine Proxy, prioritizing users who desperately needed to track local ads across different countries. While successful for that niche, it triggered a wave of negative emails from our core, long-time users.

By focusing so heavily on new, powerful enterprise features and removing the free tier, we had alienated the people who relied on the original tool's raw simplicity. Instead of ignoring the backlash, we engaged directly with these users, asking them exactly how and why the legacy tool fit into their daily workflows.

This feedback was a harsh but necessary reality check. In response, we quickly pivoted and rolled out a modernized Legacy Search MVP. This restored the soul of the original iSearchFrom—a fast, simple URL modifier—and we ensured it remained completely free. Within a week of rollout, support complaints dropped by 90% and legacy retention stabilized, successfully serving both enterprise ad-trackers and our traditional power users.

The initial MVP (left) successfully validated the need for complex ad tracking, but its layout felt disjointed and lacked the data density required for a professional SaaS environment.

For the final version (right), I completely restructured the architecture. By migrating the heavy configuration settings to a dedicated sidebar, I maximized horizontal space for the core data view. This refined, high-density UI allows enterprise users to execute rapid, precise searches without constant scrolling or context switching.

My first attempt to modernize the legacy tool (left) accidentally carried over too much structural complexity from the premium side. It felt like a stripped-down enterprise tool rather than the quick, frictionless utility our users expected.

The final design (right) corrects this by stripping away the clutter—removing heavy sidebars and nested menus to return to a clean, focused, single-purpose interface. It gives veteran users exactly what they loved about the original tool, but elevated with a modern, reliable design system.

One of the biggest learnings post-MVP was that existing users were genuinely confused by the new features. For a decade, they only knew iSearchFrom as a single-page URL modifier. Dropping them into a complex dashboard with terms like 'AI Visibility' and 'Residential Proxies' caused immediate friction.

This led directly to a major marketing strategy pivot: I needed to build dedicated feature landing pages to educate the base. Instead of a generic homepage, I designed deep-dive pages for both the Search Engine Proxy and AI Visibility Tracking.

These pages acted as educational bridges. They broke down complex concepts using clean, scannable data visualizations and targeted UI mockups.

The biggest lesson from this project was understanding exactly where AI excels and where it falls short in the design process:

1. The 'Last Mile' Problem: AI is incredible at getting you to 80% completion in record time. It can scaffold complex tables and generate mock data effortlessly. However, the final 20%—the micro-ergonomics, the visual polish, and the nuanced understanding of *how* a power user actually interacts with a data cell—still requires deep human intuition and meticulous design craft.

2. Code as a Design Tool: I learned that stepping out of Figma and using AI to generate functional UI code (like Svelte components) is often a better way to test complex interactions. It forces you to confront the reality of the browser much earlier in the process.