--- title: "When a ‘Simple’ Micro-SaaS Isn’t So Simple: The Story of TrackChanges.app" url: https://makethin.gs/posts/trackchanges-micro-saas-story author: Lyubomir Petrov published: 2025-02-22 topics: [Engineering] tags: [microsaas, spreadsheets, ai, startup-journey, alpha-testing, sabbatical, product-development, chatgpt, projects] outdated: 2026-11-11 tldr: "I set out to build a 'simple' micro-SaaS that alerts teams to meaningful spreadsheet changes. ChatGPT-generated code caught every edit but missed what humans care about, so I wrote the core logic myself and had an MVP in about 28 days. Now the plan is slow, personal growth: 1–3 alpha testers, then 10–15 hand-picked users, then marketing." --- # When a ‘Simple’ Micro-SaaS Isn’t So Simple: The Story of TrackChanges.app > **Outdated since Nov 11, 2026:** I shut TrackChanges down. A few months after launch, built-in AI capabilities covered what it did, and since it was mostly a fun project, I killed the app and the idea. The story of building it still stands. ## Motivation > **Insight:** A spreadsheet can go from ‘tiny helper’ to ‘massive headache’ faster than you’d think. I wanted to build a real, working product—just for fun. My idea was to create a _simple_ micro-SaaS (ha, more on the “simple” part later) that tackles a very common pain point I’ve seen in many companies. In my past roles, one tool always popped up: **Spreadsheets** (Excel or Google Sheets). They’re fantastic for organizing data and processes. Originally, we’d use them on our local machines—no big deal. Then came Google Sheets with its “multiplayer” aspect. But until a few years ago, even that was mostly one person entering data while everyone else just viewed it. **Fast forward to now**, and big companies with elaborate internal systems often realize they’re missing certain functionalities in their CRMs or ERPs. How do they patch these gaps? They whip up a shared spreadsheet in Excel or Sheets, letting multiple people edit away. That works fine…until you have hundreds of rows, daily changes, and no one knows who did what or when. **Ever seen a project slip three months behind, and nobody notices until the deadline passes?** The updates were in the spreadsheet, but no one had time to check it daily (or even weekly). Yes, Excel and Sheets have comparison tools, but with frequent edits, you end up digging through tons of changes just to find that one critical update from two weeks ago. That’s where the motivation for my new micro-SaaS came from. ## At First… > **Insight:** AI can write code, but you still need a human who knows what matters. I figured this would be super **simple**: let users paste a spreadsheet link, and boom—get real-time notifications about every change. Easy, right? I also wanted to experiment with AI, so I tried having ChatGPT write the code. It looked promising: it caught row additions, cell edits, you name it. So I built the marketing pages, the user interface, authentication. Then I took it for a deeper test drive—and realized it was basically _useless_ in a real-world sense. Why? Because **humans** don’t care that a formula changed from `=SUM(A1,A2)` to something else automatically. We want context—who changed what _that actually matters_ for our projects, orders, or inventory. AI alone didn’t grasp that nuance, no matter how many prompts I threw at it. So, I rolled up my sleeves and wrote the logic myself. I also tried Cursor IDE for AI-assisted coding (spoiler alert: it didn’t fix everything). This turned out way more complicated than I’d hoped, especially for a micro-SaaS. Honestly, if I didn’t have 20+ years of experience as a developer, product manager, team lead, QA, etc., there’s no way I could have pulled it off alone. After about 28 calendar days, working roughly 6 hours a day for 3–4 days a week (so maybe 9–12 full workdays total), I had a fully functional MVP. That’s actually pretty sweet, considering my laid-back sabbatical schedule. ## It’s Alive—Day 0 > **Insight:** Rushing to get thousands of signups without real feedback is like throwing a party where nobody cares about the music. So I have this MVP. I’ve tested scenarios, written automated tests, done some calculations, set everything up—and now I’m wondering, _“Where do I find my users?”_ In typical startup culture, you’d blast it on Facebook, Twitter, LinkedIn, or Product Hunt to get as many eyeballs as possible. But for a niche micro-SaaS, that can backfire by bringing in random traffic that doesn’t convert—or worse, tries it once and never comes back. My plan? Start small and personal: 1. **Alpha testing** – Identify 1–3 users who truly need this. Get direct feedback, iterate, improve. 2. **Validation** – Scale up to 10–15 users, still hand-picked. They’ll be using the _real_ product, but we’re not calling it “alpha” anymore. 3. **Next Level** – Once it’s polished, we’ll do the big marketing push. I’m still in Phase 1, and waiting for those first alpha testers can be tough (I _know_ my own impatience!), but that’s the game plan. I want real-world feedback to shape the product before going all out. ## Summary > **Insight:** Yes, I’m on sabbatical, but I still need milestones to keep me on track. I’ll stop here for now. I’m on sabbatical, working around 6 hours a day, 3 days a week (blogging included). I’ve set a goal to land at least 1 alpha tester by the end of this month, and 10–15 users by the end of March. Even with a relaxed schedule, having concrete KPIs gives me direction. Once I move beyond Phase 1, I’ll share the lessons from finding those initial users—and where I stumbled. **Thanks for reading.** If you’re curious, want to test it, or just say hi, feel free to reach out. _Stay tuned, and let’s see where this journey takes us!_ > **Note:** If you’re curious to see [TrackChanges](https://trackchanges.app/) in action, it’s live now—I’d love to hear your thoughts! Thanks for reading, and let’s keep building great things together.