Took my interview prep SaaS to $10k MRR in under a month, solo. Here is everything that worked for SEO and conversion, and what was a waste of money.
retained text
I run Prepper, an interview practice platform for trading, investment banking and consulting (disclosure: my product). It just crossed $10k MRR, barely over a week after launch, with no funding and no team. Posts like this helped me a lot early on, so here is the full breakdown. \*\*Background\*\* I spent six years in investment banking and two in private equity, and went through consulting and trading interview processes myself. The prep material was either a $500 course or a PDF. Nothing timed you and marked your answer, so I built that. \*\*The start was bad\*\* First day of paid ads: 290 visitors, zero sign-ups. People loaded the site and left. I assumed the ads were wrong. They were fine. The path into the product was the problem. \*\*Conversion: what moved the number\*\* 1. Let people use it before they sign up. Every test page now opens with a playable trial: 20 mental maths questions in 2 minutes for trading, 5 case maths questions for consulting, no account needed. You only see the sign-up when you have a score you want to keep. Visitor to sign-up went from 0% to 30% 2. Changed the popup. I opened with "50% off". Nobody wants a discount on something they have not tried. It now says "Try Prepper for free" with one-click Google sign-up, and it shows after 15 seconds or some scrolling instead of after 2 seconds. The discount only shows on the pricing page, where people are actually thinking about price. 3. Ads land on what they promise. The trading ad goes to the trading test, the consulting ad to a case maths drill. Never the homepage. 4. Welcome email with one next step, sent to Google sign-ups too. Obvious, and I had missed it for the first 100 sign-ups. \*\*SEO: three page types do almost all the work\*\* 1. "Alternatives" and "vs" pages. Things like "\[competitor\] alternatives" and "\[competitor A\] vs \[competitor B\]". People searching these are already shopping. My rule is that the pages are fair: every competitor price comes from their own site with the source linked, and each page says what they do better and what my product does not do. Fair pages rank, people trust them, and AI assistants quote them. 2. One page per practice test. Each has the format, the scoring, and the playable trial right there. 1 of these now brings in an estimated 5k visitors a month. 3. Guides that answer one question in the first sentence. "Should you guess on a negatively marked test?" gets the answer first, then the maths, then a link to practise. Roughly 500 guides in total. Also worth doing: an llms.txt file and writing so the first sentence stands alone. 20% of sign-ups now say they found me through ChatGPT or Claude. \*\*What did not work\*\* \- Sending paid traffic to the homepage. \- Adding countries because the clicks were cheap. The ad platform spent most of the budget there and I learned nothing about my real markets. \- Judging a campaign on fewer than about 150 visitors. Early on I nearly killed things that later worked. \*\*What worked better than expected\*\* Short videos of the product itself. A screen recording of the mental maths test with three lines of text on it. Around 5k views and 500 sign-ups for zero ad spend, and it took minutes to make. \*\*Numbers today\*\* \- MRR: $10k \- Paying users: 726 \- Price: $19.99 a month, free plan available \- Organic share of sign-ups: 15% \- Paid ads: 85% of sign-ups at about $5 per sign-up \- Best single page: 80-in-8 Optiver test \*\*If I started again\*\* I would build the no-account trial before spending a cent on ads, and I would write the comparison pages in week one. Happy to answer questions on any of it. Disclaimer: if you, like me, understand that code is now a commodity, and speed of launching is key, make sure to check out and try our MCP (ShipyardMCP(dot)com). It lets you connect all your stack once (ie stripe, cloudflare, anthropic), then launch the next projects in hour. It helped me launch 1 project a day, not in a week.
document fetched 2026-11-17 00:27 · canonical https://reddit.com/r/SaaS/comments/1xy32oz/took_my_interview_prep_saas_to_10k_mrr_in_under_a