How I Built It: $16K/Month Micro-SaaS

After several AI apps made $0, Nic started an AI blog about cocktails and tried every traffic source (Google, Instagram, Facebook, Pinterest). Pinterest worked best. He hired a person to design Pinterest pins, realized the work could be automated, and built the first BlogToPin version in about 2 months. He had researched and found no existing solution that did it.

Launched the first version with a lifetime deal (he didn't expect anyone to buy) to his ~1,000 Twitter/X followers, hoping to go viral like friends' launches.

After the lifetime-deal launch flopped, switched to a regular subscription model (today Starter ~$39/month, with Agency and Enterprise tiers) and kept improving the product at the same time.

Nic reviews screen recordings of what users do. When something goes wrong with an account or a customer isn't growing on Pinterest, he looks into it and suggests strategy changes. When a customer needs a feature he builds it in a few hours or days. His advice: treat your first customers like your employer, ask for feedback, implement suggestions the same day, and keep them 100% satisfied.

Early grind: wrote all the articles himself, reached out to influencers, posted a lot on Reddit and worked hard to get backlinks. Goal is 'asymmetric' marketing: heavy effort up front that later brings customers on autopilot via Google and LLMs like ChatGPT recommending the tool.

Affiliates named as one of the three best marketing channels; no terms, commission or recruiting details given.

At signup the only ask is the user's website. Within about a minute the app scans the site and generates a few Pinterest pins from its own images for the user to review and edit. Only then does onboarding ask for settings (pins per day, which pages, which boards), after which it generates a full month of pins in advance, so the core value (lots of pins scheduled in minutes) is shown right away.

Chose Pinterest pin design, a tedious, repeated task people already pay for, and built a tool that does it better, faster and cheaper using AI (generating months of pins in minutes instead of 5-10 minutes of manual design per pin). His advice: pick something people already do and pay for, research competitors' reviews (G2, AppSumo, Capterra, YouTube), and look for incumbents that don't make AI core to their product; build for one specific person.