4 Tips on How to Choose a Tech Stack for Your Project

6 min readMay 10, 2019

The app’s purpose determines what tech stack to apply. Not the other way around. Some features cannot be developed on a specific tool set. If you choose the tech stack without a strategic approach and a thorough analysis, you face the risk of the app performance gap. As a result, you’ll have to redevelop your product and pay twice.

In this article, openGeeksLab has prepared essential tips on how to choose the ideal mix for your project. We’ll guide you through the technologies we have the most experience with and those solutions that best suit your platform and meet performance requirements.

1. How a Proper Tech Stack Affects an App?

A proper tech stack makes your product user-centric by solving app users’ particular needs, increases its scalability, ensures nearly error-free and bug-free solutions.

When choosing specific tools for a particular project, you should consider, among other things:

user experience goals
a platform for building an app
a go-to-market strategy
load requirements
network conditions
on what devices an app would run.

Based on your business needs and tech requirements, you define app architecture — what languages, frameworks, libraries, patterns, servers, UI/UX solutions to combine to develop your product.

2. Most Popular Technologies

Below you can see the top languages over time according to GitHub:

Image source: GitHub

And here are the most popular technologies, frameworks, libraries, and tools according to Stack Overflow:

Image source: Stack Overflow

As you can see, JavaScript has been a leader among languages for several years now. As for frameworks, Node.js, Angular, and React are the three most popular ones.

3. Tech Stack We Use at openGeeksLab

Our main technologies for frontend are JavaScript, Angular, React.js, React Native, Vue.js. As for backend, we use Node.js, JavaScript programming software, Koa2, Express. To ensure data storage and heavy load processing, we apply AWS. To increase performance, we use various databases, depending on the tasks — MongoDB, Elasticsearch, Redis along with other tool sets.

JavaScript doesn’t give the highest speed, but it has a large community. Python and Go are faster, but the community is smaller. Therefore, the development will go slower. It is also likely that in some instances it will be necessary to develop some features from scratch instead of taking ready-made ones. If we are talking about product scalability, JavaScript is the best option. As a result, you receive a good product in a reasonably quick time.

Applying React Native development ensures creating a single app for both Android and iOS devices at once. You receive a native code, not a hybrid app by achieving the same high quality as if you build an app separately for each platform. Moreover, React Native allows shortening the development process 1.8 times, reducing costs, and ensuring quick product refresh if needed.

As Node.js is universal, you can use different architectural approaches. It proved itself as a great engine which is very flexible — using Node.js is equally effective for massive projects as well as for small ones.

It doesn’t matter which tool set to choose from. Thanks to proper architecture, it can produce excellent results. For example, in one of our latest projects GOeureka in the field of online hotel booking, we applied Node.js, micro-service architecture, and Big Data analytics which allowed us to decrease request time from 70 to 4.6 seconds with unlimited number of sources with data accuracy 95%.

4. Factors to Consider When Choosing a Technology Stack

Don’t Be a Copycat

It’s one thing to get some insights into tech solutions and experience of top brands or your competitors. Another case is to follow someone’s strategy blindly. As we’ve already mentioned, your project has a specific purpose, business logic, and requirements which determine a tech stack.

For instance, our recent project Nimbla was founded to protect SMEs from unpaid invoices. As it has a complex business logic, it required thorough research, niche analysis, and planning. Although we needed to develop it as a flexible highload web app, provide integration with third-party providers, and ensure smooth performance, a platform had to be mobile friendly and very easy to use. We applied Node.js microservices architecture to process a significant amount of requests, Angular and React, and API integration with third-party cloud accounting software with Codat.

Choose the Right Platform

Specific tools are right for building an application which runs on one particular platform, while other optimally suit for developing an app that runs on multiple devices simultaneously — iOS, Android, web platform.

React.js is perfect for developing applications with simpler logic. Angular, in turn, is a perfect match for creating apps with more complex logic.

For relatively simple solutions, you can use the PWA technology (that stands for Progressive Web Apps) which allows you to install a web application on a mobile device. With little effort, you can convert an app written in React.js, Vue.js, and Angular to a mobile application.

It’s worth considering that PWA is not able to completely replace a mobile app, but it is great as an additional option for simple application development.

Performance Requirements

Performance features vary according to the required response time and number of requests of a platform.

You can increase the speed of data output using databases. Below you can see the top databases according to Stack Overflow.

Image source: Stack Overflow

For example, you can choose between using MySQL and PostgreSQL software. PostgreSQL is faster, but you have to pay for it. At the same time, there is an alternative to using NoSQL databases.

For example, using MongoDB in some conditions is faster than PostgreSQL, but in some cases, it is slower. It means that you need to work correctly with MongoDB.

However, unequivocally, NoSQL databases allow you to store complex structured data, which is almost impossible to do in SQL databases. At the same time, NoSQL databases have quite a wide choice with their features of work.

MongoDB is the main and most universal work tool, with fast read, write, and search performances. An alternative to it is Elasticsearch technology. It has a problem with slow writing, but at the same time, it is one of the fastest in terms of reading and searching. You can also choose Redis. It is extremely fast in terms of reading and writing but has practically no data search functionality, which significantly reduces the possibility of Redis implementation.

What Tech Stack to Choose?

As you can see, finding an ideal mix isn’t a super simple task. It requires a thoughtful analysis of your business objectives, planning efforts and building software architecture. The core idea is that you should choose the toolkit under the needs of a specific project. Don’t blindly follow tech solutions proven by industry leaders. Your go-to-market strategy is of higher value than a particular toolset.

And we at openGeeksLab are the right team for you to build the right product and guide you through the tech solutions that match perfectly with your challenges.

If you have any further questions, please don’t hesitate to contact us.

Meet the Geeks:

This article was originally published on the openGeeksLab blog.