SaaSJanuary 31, 2026

30 Days of Building a SaaS — I Solved 54 Problems

From validating ideas to deploying live, I tackled 54 common indie dev problems in 30 days. Demand, tech, payments, operations — a real daily log of every obstacle and breakthrough.

01

My first project was a one-time payment product. Two days ago I also finished building subscription-based payments and wrapped everything into reusable modules — though this project doesn't need subscriptions. It's for the next project, and the one after that, so I can just plug and play.

With that, my first project is officially complete.

So what exactly did I learn? What problems did I solve?

I think it's worth taking inventory, because honestly, I've lost count. This isn't a humble brag — it's real. Let's look together.

02

2026-01-01

  1. When all 10 ideas I was validating led nowhere and I got stuck in a loop — what do I do? Here's what I wrote in my journal.

Yes, lower the bar. Does the pain point need to be severe? High frequency? Large audience? Strong willingness to pay?

No. For the first project — mild pain is fine, low frequency is fine, zero users is fine, no willingness to pay is fine.

Lower every condition. And that's how this article was born.

2026-01-02

  1. Look, I fell back into the "it won't make money" trap again. Had to correct course, hah. I instantly pulled myself back to the purpose of this round: start the process as fast as possible.

  2. Solved the domestic vs. overseas market question — and decisively chose to go global.

  3. Worried about getting stuck on unsolvable problems during development and not knowing how to ask for help? That's how this article came about.

2026-01-03

This phase solved these questions:

  1. What if I don't understand overseas products?

  2. Should I even bother learning about overseas products?

  3. Should I tackle something harder?

Also figured out how to find overseas ideas.

My first project idea actually came from AI, hah.

I also noticed Cursor was too slow — but no rush, that'd get solved later.

2026-01-04

  1. Learned how to validate demand. It's an incredibly tedious, repetitive process — searching the same keyword over and over in different variations. That's right, valuable work is like this. Stay patient.

  2. Solved the "frontend dev who can't do backend" problem. As a complete beginner, with AI's guidance, I learned to configure Supabase, create tables, and fetch data from the frontend. Confidence slowly built up.

2026-01-05

  1. Learned how to choose and purchase AI coding tools. Back then I thought using a proxy was the most convenient option. Now I believe the best approach is buying an official CC Pro membership through a clean home IP + a reliable reseller.

  2. Learned how to use Claude Code. Don't laugh — seriously. Since finishing a mini-program with Cursor in early 2025, I hadn't touched any coding tools. I literally followed a tutorial step by step just to learn how to type claude to start it.

2026-01-06

  1. Learned how to get Claude to teach me prompting for Claude Code — completed the UI and basic business logic.

  2. Solved the "text getting swallowed" issue when inputting long text in Claude Code. I almost cleared everything and retyped it character by character, then asked Claude and discovered it was just collapsed.

2026-01-07

  1. Solved the problem of bringing business context into a new conversation after being forced to restart due to long chats with AI.

  2. Solved the conversation flow problem with AI. Don't bring up development until the requirements are fully understood. Once you jump into development planning, you'll write tons of redundant code. It's the same as at work — 9 years as a programmer built invisible but invaluable experience.

2026-01-08

  1. Feeling like I can't keep going, so frustrated — what do I do? Write it down. See what's actually causing it.

  2. I clearly did a lot during the day, but at night it still feels like I did nothing — what do I do? Review the day's TODO list, see what was accomplished, mentally close out the day, then go to sleep. Yes, health matters too.

  3. The writing question — should I publish today? After 5 PM, if there's no material, don't publish. Health is important.

  4. I've been publishing every day for so long, but not publishing makes me unhappy — what do I do? Ah, desires, hah. But my gentle side won.

So on the 8th, I didn't publish an article.

2026-01-09

  1. How exactly should I run a WeChat Official Account? How to make it valuable? How to keep it valuable over time? After writing this down, I had my eureka moment and saw things much more clearly.

Yesterday, after I shared an article in a community group, someone said they followed me because I'm pretty, and suggested I post more selfies.

I immediately shut that down and said directly, "Then you're not my target audience."

Followers attracted by looks are obviously not valuable. Besides, constantly posting selfies isn't dignified — it mostly comes across as seeking attention. More importantly, I don't intend to be a beauty blogger. I don't rely on looks for a living, right?

So what's actually valuable? Continuously solving problems and providing value. Making yourself needed, and letting more people see and benefit from your writing.

2026-01-10

  1. Solved the domain purchasing problem. Tried registering with Wise, but verification kept failing. Tried PayPal, but it didn't work either. Ended up buying through a Taobao proxy — about 10 RMB above the original price.

  2. Learned how to set up a support email.

2026-01-11

  1. Learned how to configure a domain on Vercel.

  2. Learned how to cancel a domain subscription.

  3. Learned how to pick a safe domain name and what to watch out for. Using names related to big brands is a red line — never cross it.

2026-01-12

  1. Learned how to create a logo and convert it into favicons for every use case in one go. Just had AI generate the prompt based on the project background, then used a design tool to create it.

  2. Realized the footer isn't simple — it's a place to build trust.

2026-01-13

  1. Solved the project workflow order problem. Price first, then design the visuals.

2026-01-14

  1. Learned how to create multi-dimensional pricing for a product with low usage frequency.

2026-01-15

  1. Solved how to integrate Google login into the project. Before writing this solution, I'd also tried next-auth, but wasn't satisfied with the result.

  2. What to do when you're not happy with the result? Find benchmarks. See which competitor's solution feels the smoothest and best to use, then reference their approach.

Remember this mantra I mentioned?

Every problem I encounter, 100%, has an answer.

2026-01-16

  1. Learned how to apply for a business license. But later, since individual accounts became available, I dropped this.

  2. Learned how to calculate PayPal's actual take-home amount and which plan is more cost-effective.

2026-01-17

  1. Solved the "should I create a community group" question. No. I'm just starting out — I need to preserve my energy for building.

  2. Learned the anatomy of a SaaS homepage.

2026-01-18

  1. Learned how to open a personal PayPal business account.

  2. Learned how to remove the proxy system configuration and install official CC Pro.

2026-01-19

  1. Learned how to integrate PayPal payments.

2026-01-20

  1. Learned how to write a privacy policy.

  2. Learned how to write terms of service.

2026-01-21

  1. Learned how to deploy and go live.

2026-01-22

  1. Learned how to keep code from becoming a nightmare. For tech stack choices, pick what's long-term and stable, not what's easiest to implement. If something can be reused, don't write it twice.

  2. Learned how to do SEO for a project. Though I later realized that during the initial validation phase of a SaaS, you only need two things: deploy Google Analytics and basic meta tags. Everything else can wait.

2026-01-24

  1. Solved the "what to do when your live project has an emergency" problem. A 404 page is too blunt — what's a better approach?

2026-01-25

  1. Understood the anatomy and importance of OG (Open Graph) share images, and learned how to create them.

  2. Learned about the stages a SaaS product goes through after launch — also helped me understand when to abandon a project.

  3. Time management conflicts causing anxiety — what do I do?

2026-01-26

  1. Learned how to integrate Creem's one-time payments.

2026-01-27

  1. Solved the sitemap auto-sync problem. Manually updating it after every file change was too tedious — this had to be automated, and it'll save a ton of effort for future projects.

  2. Added a feedback section to the project.

2026-01-28

  1. Solved how to centralize all project feedback in one place.

2026-01-29

  1. Learned how to integrate subscription payments for both Creem and PayPal. PayPal was slightly more complex than Creem.

2026-01-30

  1. Recognized the importance of managing overseas social accounts. They're a bridge for building trust with international users — time to take this seriously.

03

Of course, I didn't mention the specific technical bugs and issues during development — those aren't worth listing here. I only extracted the common problems that every indie developer faces.

Looking back at this month, it was really just non-stop problem solving.

Some problems seem simple now, but I was stuck on them for ages back then. Domain configuration, payment integration, OG image setup — each one had to be chewed through bit by bit.

The good news is the foundation is built now. Future projects will be much faster. Maybe 10 days, 8 days, or even a single weekend.

This isn't talent — it's built from time and problems.

But I'm also clear that this month was mainly skill accumulation. The real challenge lies ahead — how to make a product that users genuinely need, and how to continuously create value.

Take it slow. Let's keep going, together.