I built 104 free tools solo in 4 months. Here's the honest journey so far.
I'm Talha, a solo developer from Pakistan. Four months ago I got tired of every "free" online tool site being covered in ads, popups, and forced signups just to compress an image or merge a PDF. So I built my own.
This is the real story, what worked, what didn't, and where things stand right now, numbers included.
What I built
DukoTools 104 free tools across PDF, image, developer utilities, finance, text, and a few niches nobody else covers: a Zakat calculator, Pakistani electricity bill (WAPDA) calculator, CNIC validator, cricket run-rate calculator, and construction estimating tools.
No ads. No signup. Most tools run entirely client-side in your browser, nothing gets uploaded to a server unless the tool genuinely needs backend processing (PDF conversion, video downloading, a couple of image tools).
Stack: Next.js frontend on Vercel, FastAPI backend on Railway, 5 languages (English, Spanish, Arabic, French, Urdu).
The numbers, honestly
This is the part most build-in-public posts gloss over, so I'll just say it plainly:
- 4 months of building, mostly nights and weekends
- Peaked at 76 active users after a marketing push across Reddit, Dev.to, Product Hunt, and Hashnode
- Dropped back down to around 26 active users a week later, once the initial post-launch traffic faded
- 0 paying customers, the whole thing is free, monetization is AdSense, currently switched off until there's enough traffic to justify it
- 2 Product Hunt launches so far, both technically "the same product," which I now realize was a mistake, PH doesn't reward relaunching the same thing twice
What actually went wrong
Reddit banned my account. I posted twice, promotionally, from a 2-year-old account that had zero karma. Didn't matter that the account was old, zero karma plus a promotional link reads exactly like a bought/hacked account to Reddit's spam detection, and it got flagged.
Then I made almost the identical mistake on Quora. New account, and I spent a day answering a bunch of genuinely helpful questions, PDF conversion, OCR, Zakat calculations, mentioning my site naturally where it was relevant. Every individual answer was solid. But answering too many similar-topic questions in one day, each pointing at the same domain, read as a spam pattern to their automated detection anyway. Banned. Same root cause as Reddit, different platform: volume and repetition trigger these systems regardless of how genuinely useful the content is.
I killed a tool because of cost, and it cost me more than I saved. My background remover was burning through backend compute faster than expected, so I pulled it from the frontend without setting up a redirect. What I didn't realize until users started dropping: that tool was specifically called out in my Product Hunt launch post as a differentiator. Anyone landing on that old URL just hit a 404 and bounced.
The pattern across both banned accounts is the same lesson, twice: trust has to be built before promotion, not alongside it, regardless of how good the individual content is.
What's working
- SEO/indexing is actually done now. Sitemap submitted and confirmed indexed on Google and Bing (verified through Bing's URL Inspection tool directly). Yandex is processing. IndexNow is wired into the deploy flow, so new tools get pushed to Bing/Yandex instantly instead of waiting on a scheduled crawl, one API call, no more manual submission.
- Dev.to has been the most sustainable channel by far. Technical deep-dive posts about actual engineering problems, a circuit-breaker pattern in the video downloader, memory/decompression-bomb handling in the PDF tools, get real engagement. Pure "check out my tool" posts don't. No account restrictions here either, unlike Reddit/Quora.
- The niche tools (Zakat calculator, CNIC validator, cricket tools) are genuinely differentiated, I haven't found another "everything tool site" that covers this ground.
What I'm doing next
- Rebuilding Reddit from a fresh account, genuine, link-free participation for 3-4 weeks minimum before ever mentioning the site, spread across unrelated subreddits so it doesn't just look like a marketing account
- Not going back to Quora yet, appealing the ban, but not counting on it
- Fixing the redirect gap left by the background-remover removal
- Setting up proper GA4 key events so the next round of decisions is based on actual tool usage, not vibes
- Launching my South Asian tools collection as its own standalone Product Hunt listing, rather than relaunching the whole site a third time
No polished conclusion here, I'm genuinely still learning as I go. If anyone's navigated the "genuinely helpful content still gets flagged as spam" problem on Reddit or Quora specifically, or has thoughts on when it's actually worth turning on AdSense versus waiting for more traffic, I'd like to hear it.
Site's here if you want to poke around: dukotools.com
The sheer volume of 104 tools in four months is impressive, but the real play here seems to be your focus on local-specific utilities like the Zakat and WAPDA calculators. Most developers chase generic global tools, but by solving highly specific regional pain points, you've built a defensible moat that a generic PDF converter can't easily replicate. Since you are running these client-side, have you tracked if users are returning to specific pages as recurring utility hubs? It might be worth looking at your logs to see which of those niche tools acts as your primary 'anchor' to bring in organic search traffic versus which ones get one-off usage. Focusing your maintenance and future build efforts on those high-intent categories rather than expanding the total count might offer better long-term compounding.
I think your biggest takeaway is bigger than Reddit or Quora. You built 104 tools, but your niche tools are the ones people actually remember. I'd double down on those instead of trying to compete as another "everything tools" site. Distribution matters, but a clear reason to exist matters even more.
The GA4 point stood out — especially after four months of learning mostly from traffic and channel-level signals.
Once you can see actual tool usage, what decision are you most hoping that evidence will let you make with confidence?
The 76 → 26 drop isn't a failure, it's the actual signal. 76 was the launch spike, 26 is the number of people who had a reason to come back. A spike tells you distribution worked; the floor tells you whether a tool earns a repeat visit. Before turning AdSense on, I'd track one number: % of users who return within 7 days to a tool that isn't part of the launch push. One returning loop beats 104 one-off tools.
That reframe is exactly right, and it's uncomfortable in a useful way, I've been treating the drop as a problem to explain rather than a number to actually act on.
The 7-day return metric is the one I don't have right now, and reading your comment made me realize why: I've been tracking raw active-user counts in GA4 but haven't set up key events yet, so I genuinely can't answer "which tools bring people back" versus "which tools got a one-time visitor from a launch post." That's the actual gap, not the drop itself.
Going to set that up before touching AdSense at all. If the honest number is "3 tools have a real returning loop and 103 are one-off utility visits," that changes what I build next far more than the total user count ever would.