Loading
Founders and technology leaders often face a tempting offer: “app templates” or “app accelerators” that promise faster builds and lower costs. While these solutions might seem appealing at first glance, it’s crucial to understand their true impact on your project’s long-term success. This article discusses templates and helps you navigate the topic when building an app.
The offer might sound efficient, but templates often distract from what you really need and result in more mess when you release the app. Moreover, they can simply mask overcomplicated design in the first place. Nimblesite understands this better than most. We focus on a simpler, more sustainable approach. We don’t use complicated engineering patterns and then pretend to give you a head start with an “app template” or “app accelerator”. Let’s explore why.
App templates or accelerators are preexisting cookie cutters for apps or systems. They usually come with some existing UI flows, like sign-in and registration or in-app purchases functionality. A company that builds apps may reuse these templates again and again for multiple clients. The idea is to get moving faster at the beginning.
Templates look like a quick solution. They come packaged as polished frameworks that claim to cut corners and costs. The pitch: “We’ve done half the work so that you can start closer to the finish line.” A template may help the developer get you the first version of the app faster, but this is largely redundant now that AI coding assistants can help with this process.
If your developer is building a gigantic, overcomplicated system that doesn’t meet your direct business needs, they will need what we call “boilerplate”. Boilerplate is more or less the fluff that developers spend time on that is not entirely necessary. It’s there because they’re following an approach that they heard is a good idea or because it’s the only approach they know.
Seasoned developers minimise boilerplate, but more cynical companies will pile it on and charge you for it. That’s where templates come in. They give you the boilerplate that you shouldn’t need in the first place.
Many consultancies push heavy frameworks like Clean Architecture and Bloc because they pad their value proposition. It makes it look like they’ve solved a hard problem and provided structure. But the truth is that boilerplate code isn’t a real problem. It’s an easy thing that AI can handle in a snap. The real challenge lies in the business logic and the product’s differentiation.
These so-called accelerators are often marketing ploys. They make it look like you save time and money, but they really just deliver a bundle of unnecessary code that may have no value to your unique application.
You want less code. Code costs time and money to maintain, and all the code in your systems needs to be upgraded and tested over time. More code means more complexity for the developers who work on your app. So, all things being equal, you should never add unnecessary code to your system. Yet app templates do precisely that: pack tonnes of unused code into your app before you even start.
They pack in layers of unused features or “handy” helpers that never get touched. This leads to more cleanup, more overhead, and more wasted time. The “shortcut” ends up costing more in the long run.
“All code is a liability” applies here. Every extra line of code adds complexity and a potential cost down the road. Instead of starting with a bulky template that you must trim back, you should focus on small, carefully chosen libraries. When you pick low-impact libraries, you start lean and prevent work in the future.
Suppose you add only what your product needs. You avoid the heavy lifting of throwing out giant chunks of code you never needed in the first place.
Firstly, your developer needs to simplify the design so that most boilerplate is unnecessary. But when small amounts of boilerplate are necessary, AI coding assistants come to the rescue.
AI coding assistants have changed the game. They can generate boilerplate code on demand, and they adapt to your coding style and strictness. They fit in neatly with tight coding rules. They don’t jam your project with pre-built templates that never quite match your exact specs. Instead, these assistants do what templates claim to do, but they do it custom-tailored to your needs. They help fill the small gaps without forcing you to take on a giant pre-made pattern.
The best part is that they don’t generate more boilerplate than you need, and they can clean up after themselves when you finish.
See our article on building Flutter apps with ChatGPT.
Check out Nimblesite’s minimal pattern, Simplified Unidirectional Data Flow. We recommend it as a good compromise between robust architecture and minimal boilerplate.
Of course, many people are sceptical of AI. It often hallucinates and makes mistakes. This is where the role of automated testing is critical. It would be foolish to accept code from AI at face value.
Nimblesite prioritises automated testing in our quality process. We use various quality gates, such as static code analysers, but ultimately, we ensure that your app works by testing it. AI can help even with this, but there is no substitute for making sure that the user can traverse all the main workflows and that these don’t get broken on every new release.
Leaders should demand direct, lean approaches to building apps. Stick to the basic building blocks. Skip the templates. Embrace small libraries and AI-driven shortcuts when you need them. Avoid any approach that starts with a pile of boilerplate and fluff.
That doesn’t mean you skimp on quality. Instead, use the time and money you save to invest in automated testing, which is the single largest predictor of quality for your app.
Your product, your team, and your bottom line will thank you.