ShipFast: Your Fast Track to Full-Stack Web Mastery!

You come up with what seems like a great idea for a startup. You sleep on it and it still sounds good, so you come up with a nice domain name and buy it. You’re excited to start building!

But before you do, you need to decide on a framework. You spend 1 hour looking at what most people are using and decide on NextJS. It’s well supported, lots of people use it, has good docs.

You run npx create-next-app@latest, tell it to install TS, ESLint, Tailwind, App Router, a few other configs and get going.

You start by building the public and marketing pages you’ve mocked up, to present your new thing to the world.

But you quickly realize you’re gonna need email marketing. It’s not installed yet, and you know you’ll want to research the best Marketing Email Providers with a generous free tier, so you leave it for after you’re done with the public and marketing pages. It might cost you a few subscribers, but no big deal.

Work, work, work. Ok, the public and marketing pages are done.

A bit of a frustrating and tiring experience because, even though these were mostly static UI components, the UI library you're using is a little too low-level and mixes design and functionality, forcing you to mentally switch context between the two.

But it's done. You're tired but happy! Time to integrate with an Email Marketing Provider.

You spend 4 hours researching the best ones, decide on one with a generous free tier, and spend 1 hour creating a beautiful newsletter component that sends the user’s email address to the backend.

Now that you have the user’s email on the backend you ask yourself if you want single or double opt in. You know you’re less likely to receive spam complaints from users receiving your emails with double opt in.

But it’s more complicated, and you’re tired - and this is just a startup anyway. Single opt in it is.

You go to sleep feeling proud of having finished the public and marketing part of your app.

A new day dawns. You’re feeling motivated to continue building your new project.

But it’s time to think about the hardcore stuff: database, data model, authentication, permissions, user flows, business logic, deployment pipeline 😮‍💨 ufff this is going to take some work... You take a deep breath and get going.

Work, work, work.

Two days have passed. Your went for MongoDB and Mongoose.

The schema is ready to go. Basic authentication with registration and login is working. Time for protected routes, profile page, email and password change flows. Another deep breath... let's go!

Work, work, work.

Two more days have passed.

Authentication is done and you're proud. Registration, login, and email and password changes are working.

There's no Two-Factor Authentication or Single-Sign On, but this is a startup anyway. We'll do that later if it succeeds.

Ok, you're almost ready for the interesting stuff: building the actual app.

But before you do, you need to get this thing online. So you spend 2 more hours researching the best CI/CD techniques and which features each cloud provider offers.

Two more days have passed.

You now have a CI/CD pipeline you're feeling proud of, deploying on Vercel servers. It's quick and easy.

A week has passed since you've started and you're now ready to start building your app.

Finally! With an incredible amount of sustained excitement, you're eager to get started.

You build your app, show it to your users and, incredibly, it succeeds! People actually start using it, amazing! Dreams of profitability start to materialize and it's time to install Analytics. This looks like a no-brainer decision, use what everyone else uses.

So you install Google Analytics. And a cookie banner to comply with GDPR.

You're on a roll, happily pushing new features until...

Uh oh, user activity starts to drop abruptly :/ What the hell's going on? Damn it, you probably have an error that's breaking the app. You spend 2 hours frantically searching for the bug but you can't find it so you need to install a monitoring tool that supports session replay urgently before enough users leave and you have no failed sessions to debug.

Work, work, work. 2 hours later you push a commit with the monitoring tool's code!

There's still a few users left, so you can do a session replay and find the exact sequence of actions that lead to the bug and correct it! Ufff 😮‍💨 crisis averted 😅

You send an apology email to your users and they're happy to see you're responsive.

Would have been better if this bug was caught early, but hey, could have been worse.

The same goes for uptime issues, performance issues, memory leaks, SEO, automated testing, linting, type checking, developer experience and so many other aspects of development that are not directly related to actually building your stuff... And this process is almost the same every. single. time.