Most project templates are either too simple or absurdly over-engineered. Here's what actually works for building MVPs fast.

I'm a developer who has worked with many clients from multiple domains and countries. So I used a lot of templates to boost up my speed. And from those usage, today I'm sharing the key pain points I've discovered while using templates around the internet.

1. Over-engineering

It's wonderful for a template that already cover the future scaling if the client needed. But sometimes, it adds unnecessary complexity to the template itself. Example, if I want a micro-saas template that ready to give me to MVP ASAP. Setting up a project with k8s, microservices, RabitMQ is not only time consuming, energy drawing but also being overkill. Even though that is a good stack for when going big, but for a micro-saas with target of getting MVP ASAP to have a research about the market? It's a big down side. With me, a good template is cover scaling ability, but recognize the threshold—knowing when good enough is actually good enough to stop scaling more. A monolith project with base fundamental covering that good enough so project not getting messy to quickly, but not a gigantic walking machine that handle 100k concurrent requests is a sweet spot for me.

2. No test

Untested templates are inexcusable. What's worse for a client when they receive the thing they bought just to find it in an unusable condition? This is the absolute bare minimum of an item before reaching the client, no one is willing to buy a piece of junk after all.

3. English only

It's true that English is used widely all over the world, but there are still other languages too. Making a template for a globalized language is a good thing, but with me it's kinda not enough. Just by simply adding support to one other language at least, you have easily unlocked much more the size of the client pool. From browsing marketplace sites like Gumroads, Colorlib, Envato,... I've noticed multilingual templates consistently have higher sales than templates that only support one language, and base on that we can tell that we - human really want somethings that could help us getting things done faster but requiring less effort.

4. Too many friction

We all want things that go swiftly, no one wants to have something that go slow and wasting time to wait for it to run. And the speed I'm talking about right here is the time taken of client to use the product they bought. When a product reaches a client, they may still have to wait for hours for it to build? That's a major red flag. Instead of that, a product that starts with one command, and also boots up in seconds is a perfect thing that everyone wanted.

5. Bloated

A template with such simple built in function but taking too much space is a dead product. We have evolved for years, turning a gigantic computer that a size of a building now just fit on a desk, so why we have a template that bloated like a couple of GB just for very few functions right? Compact it, de-bloat it is a must to do for better user experiences

With me, sometimes the best template is the simplest one. No need to go crazy, no need to be huge. Just covering the base fundamental things that everyone wants to avoid is the best.