Why I Built LeadHunt: A Founder Story
I was paying Apollo $200/month and still not getting what I needed for my agency clients. So I spent 8 months building the tool I wish existed when I started. This is the story.
The frustration
In 2024 I was running AgentieSEO.net, a marketing agency in Bucharest with 6 people. Half our revenue came from local businesses — dentists, clinics, law firms, restaurants. The other half came from tech companies. The local half required 3x more work to acquire and 2x less revenue per client.
The painful part was the lead gen. I was paying Apollo $200/month. It found 30% of the local businesses I needed. For the other 70%, I was manually Googling, copying to spreadsheets, and running website checks by hand. 8 hours a week on something that should have taken 20 minutes.
The evaluation
I spent 3 months testing every lead gen tool I could find: Apollo, Hunter, Lusha, Snov, Skrapp, UpLead, Clay, Phantombuster, Seamless, Lead411, Cognism. 12 tools, 12 free trials, 12 disappointments.
The pattern was always the same: great for enterprise B2B, weak for local. Most did not have website audit. None had WhatsApp. None had CRM. To get everything I needed I would have to pay 3 different tools totaling $300+ per month and still do the audit step manually.
The build
In August 2025 I opened a new Python project folder. First commit was a 40-line script that called Google Places API and returned 20 dentists in Bucharest. That was day 1.
By month 2 I had the audit checker (29 points). By month 3 I had a Flask dashboard because my colleague Catalin kept asking to run the scripts and did not want to use the terminal. By month 5 I had a CRM because I needed to track deals. By month 6 I added email and WhatsApp outreach. By month 7 I added Retell AI for voice calls.
The whole thing was built on nights and weekends. I would finish client work at 7pm, have dinner, then code from 9pm to 1am. Saturdays were 6-hour coding sessions. Around month 6 my wife told me I looked tired and should stop. I took a week off and then went back because the tool was working.
The "wait people want this" moment
Around month 6 I demoed the tool to an agency owner friend from Cluj. He used it for 2 hours and said "how much?" I had not thought about pricing. I said $50/month figuring he would laugh. He wrote me a check for $500 for a year upfront on the spot.
The next week another friend. The week after, a stranger from Indie Hackers. By month 8 I had 12 paying customers without ever shipping a landing page. That is when I knew I had to ship this properly.
What LeadHunt is now
LeadHunt is what an agency owner builds when they are sick of paying for tools that do not do the job. Every feature exists because I needed it first. Every pricing decision is based on what I would pay, not what I could charge.
- Free tier: 2 audits + 50 leads, no credit card (because people hate credit cards for "free")
- $49/mo Starter (because Apollo was $59 and I wanted to undercut cleanly)
- CRM included (because Apollo charges $79/mo extra and that is ridiculous)
- Self-host CLI option (because some agencies want to keep data on-prem)
- Public roadmap (because I hate black-box SaaS)
- Discord server (because I want to talk to users, not hide behind a ticketing system)
Frequently asked questions
Why build a lead gen tool from scratch?
Every existing tool made a tradeoff I was not willing to accept: Apollo was too expensive and missing local businesses. Hunter was email-only. Lusha was credit-based. Clay was enterprise. None of them had a website audit. After evaluating 12 tools for 3 months, the answer became clear: build it myself.
How long did it take to build LeadHunt?
8 months from first commit to first paying customer. Built on nights and weekends while running my agency. Initial version was a Python CLI that scraped Google Maps + ran audits. The dashboard came 3 months in when my co-worker asked for it. The CRM came 5 months in when I needed to track my own deals.
What is the tech stack?
Python 3.11 + Flask for the API and dashboard. Supabase Postgres for data. Next.js 15 for the marketing site. Render for hosting the backend. Vercel for the frontend. Google Places API for discovery. Retell AI for voice calls. Stripe for payments. Total monthly infra cost: ~$30.
Why release it as a SaaS?
Because agency owners kept asking for it. I showed the tool to 5 agency friends in Bucharest, they all wanted it. Then 5 became 15. Then people started asking to pay. At that point refusing to ship felt wrong.
Are you still running the agency?
Yes. LeadHunt is still the tool I use every day for AgentieSEO client work. Dogfooding is the only way I know how to build software. The features you see in LeadHunt are features I needed first, not features that sounded good in a pitch deck.