Overview
In recent years, low-code tools have become a staple in many businesses. Companies like Retool, Mendix, and OutSystems have developed useful tools, which enable software developers to craft web applications with minimal hand coding.
The fundamental idea behind low-code development is using non-code tools, often with visual, drag-and-drop interfaces, to create applications. These interfaces include different settings that can influence the layout or logic of the components being built — again, without requiring the use of code. However, in contrast to no-code tools, low-code tools allow additional customization through code when the interface reaches its limits. This code customization can happen via scripts, such as Google App Script in Google Sheets; injecting external code; or adding custom code snippets. For a deeper understanding about Low Code and what differentiates it from Code or No Code, you can read more about it here.
In this article, I’ll talk about the history of low-code development, what differentiates it from code and no-code development, and how it became the force it is today.
Rapid Application Development (RAD): 1980s and 1990s
While the term “low-code” was coined by Forrester Research in 2014, the origins of low-code tools are much older. Most experts agree that the earliest low-code tools appeared in the 1980s and 1990s, with the emergence of rapid application development (RAD) tools like Borland’s Delphi, PowerBuilder, and Microsoft’s Visual Basic. Interestingly enough, my own first foray into traditional programming was with both Delphi and Visual Basic back in the day.
The RAD methodology was conceived in the 1970s but first officially introduced by James Martin in 1991, who wanted to establish a new standard for software development by moving away from the traditional waterfall development approach and toward a more iterative and prototype-focused one. Before that time, software development projects were usually defined in advance in meticulous detail, a stage could be started only if the previous stage had been completed, and changes were made only if grave errors were encountered. RAD was a more cyclical approach, allowing developer teams to revisit earlier stages and make changes based on user feedback. In addition, RAD allowed applications to be developed in modules that could be built and tested independently.
While the RAD approach is not the current standard of software development, many of today’s best practices are closely aligned with the philosophy. Low-code tools were a logical consequence of the RAD approach, since they enabled faster development, offered prebuilt components, and often had drag-and-drop interfaces that allowed users to quickly cobble together working applications.
Model-Driven Development: Early 2000s
Model-driven development (MDD) was the next step after RAD in the low-code world. MDD is a software-development methodology that relies on graphical user interfaces to represent code’s functionality. One of MDD’s advantages is that it is often platform agnostic and enables the creation of code for many different applications. Manipulating the visual representation changes the code, so less in-depth coding knowledge is needed and more focus can be put on logic and solving the problem. Some commercial pioneers in this space include Mendix, OutSystems, and Appian. Other products included IBM’s Rational Software Modeler (love the name) and the Magic xpa Low-Code Platform. While RAD and MDD share similarities in the sense that they are intended to make software creation more efficient, they are different methodologies. RAD involves rapid prototyping, fast iterations, and reusable components, and MDD relies heavily on the visual nature of the representation and an efficient way of creating code from that. The core of both RAD and MDD can be seen in many low-code platforms today.
The Emergence of Low-Code Platforms: 2010s
As mentioned above, the term “low code” was first seen in 2014, in a report written by Clay Richardson and John Rymer.
The authors keenly observed that many firms increasingly relied on low-code application platforms to accelerate the delivery time of new programs. They often started testing the platforms for internal use and were surprised with the speed at which they could churn out new applications. They then quickly realized that externally facing products could also benefit from the infrastructure and reusable components that these platforms brought to the table. Some of the most prominent platforms during this time were made by older companies that were mentioned above: Appian, OutSystems, and Mendix. They gained an incredible number of customers, and their market value rose sharply. OutSystems is now worth 9.5 billion U.S. dollars, and Mendix was bought by Siemens for 700 million dollars in 2018.
One commonality of many of the applications built during this time was the fact that they were layers on top of the databases that stored business data from the relevant companies. One good explanation I use is that low-code platforms of the 2010s focused on the fast creation of user interfaces that could read and manipulate state in a database. This simple explanation covers various use cases, from customer support to HR, finance, and supply chain management. Even today, many low-code applications are trying to facilitate exactly that. However, other successful examples exist; the website-focused low-code player Webflow is one of them. While Webflow could technically also be classified as low code instead of no code, it does have some coding elements that bring it to the highest level, and it was originally built for professional developers. Webflow was launched in 2013 and has millions of users today.
Yet something else was stirring behind the scenes during the rise of low-code platforms. We started to see a new type of creative person, now known as the citizen developer. These folks started creating their very own business applications. Thanks to low-code application development platforms, the line between the IT crowd and the rest of the business got fuzzy — and it revved up the engines of digital transformation. These platforms essentially handed citizen developers the keys to the development environment. They found that you didn't have to be a full-time developer to build something functional. With visual development tools, a handy drag-and-drop interface, and a bunch of prebuilt components, they were able to build functional and scalable apps in no time. This left traditional developers with more time to dig into the big stuff, the real head-scratchers. And the outcome? A faster, more inclusive way of creating apps that transformed business processes and improved the customer experience. At the same time, many full-time developers started using low-code tools to build internal tools or extensions to the existing code base and were able to produce value at a much higher speed than before.
Mainstream Adoption: Late 2010s and Early 2020s
Low-code platforms grew rapidly in the late 2010s and early 2020s. The demand for software applications continued to grow exponentially, and the number of skilled developers grew at a much slower rate. Therefore, experienced developers continuously looked for ways to speed up time to value, often relying on prebuilt modules and platforms to build applications more rapidly and, almost more importantly, directly where users would interact with them. Some of the products at the forefront of that movement are Google’s AppSheet, which brought scripts to millions of spreadsheet users on the web; Microsoft Power Apps, which is powering thousands of enterprise applications; and Salesforce’s Lightning Platform, which is the foundation of apps across a wide range of business functions. A notable player from the startup scene that took the industry by storm is Retool. Its makers call Retool the platform for internal tools, and they used a refreshed design and strong distribution engine to become a billion-dollar company a couple of years after launching in 2017. Retool’s success has prompted the emergence of a whole set of companies focusing specifically on building front-ends for internal tools through low code and no code.
The Future of Low Code
Where will low code’s journey go in the future? New players, new products from old luminaries, and new technologies promise an exciting decade ahead for low-code platforms. The continued move of computing power into the cloud, the emergence of AI (and specifically transformer models and LLMs), and continued investments by both large and small companies are the foundation of a new generation of products that further simplify the lives of many working people. Software is eating the world, and how we interact and build software is changing rapidly. At Fastgen, we hope that experienced developers will become significantly more productive, and that inexperienced builders will get tools that allow them to build products they couldn’t have imagined before.
It will be a wild ride, and we’re here for it.
