The low-code category is estimated to grow substantially in the coming years. By 2024, Gartner projects 75% of large enterprises will have adopted four low-code or no-code tools for both IT application and citizen development needs. But what exactly is low-code, and what are the scenarios in which it is needed?
Low-code, like so much other software jargon, is somewhat ambiguous. Under this umbrella term exists a well of various tool types—website generators, form builders, API connectors, database builders, workflow automators and more, each requiring varying levels of technical proficiency. Some are 100% codeless and UI-based with drag-and-drop templates. Others need programming acumen to set up to complement a developer’s pre-existing workflow. When we talk about low-code, we’re typically referring to platforms that help create internal business utilities.
I recently connected with Divyesh Kharade, co-founder and CEO at DronaHQ, to demystify the low-code category. Below, we define citizen developers, break apart low-code versus no-code and dig into specific types of low-code scenarios. Given this knowledge, technology leaders can better identify what tools they need to increase rapid development for their business needs.
Hard Code Is to No-Code as Manual Is to Driverless
So what’s the difference between traditional coding, low-code and no-code? To demonstrate the nuances between emerging code paradigms, Kharade compares them to automobile types.
Manual Coding
Traditional coding is like a manual car with a gearbox
With a manual car, you must know how to switch between gears. Not to mention, you must understand driving itself and how to navigate busy streets, all while following traffic laws. We can compare manual driving to traditional coding. Manual coding is for trained developers.
For example, consider we need to develop an application to interact with a database. To create this from scratch, this could mean coding a PHP application, connecting it to a back-end database management system such as PostgreSQL and hard-coding commands for ingesting data and manipulating entries.
Manual coding requires an adept understanding of programming languages and development frameworks and libraries and collaborative development processes such as git. Naturally, traditional coding is in the professional developers’ realm and often necessary for software teams developing proprietary software-as-a-service offerings. It affords ultimate customizability.
Low-Code
Low-code is like an automatic car
With an automatic car, you must still know how to drive, but gear-switching is automated. Cruise control is a nice perk, too. Similarly, low-code seeks to make development easier. It even enables some non-tech folks to construct applications. Low-code is used by professional developers and citizen developers.
If we consider our database example, a low-code solution would likely involve a one-time registration of a database. This would require a little technical expertise to initiate, Kharade said. However, after which, it would be smooth riding—a user interface would be utilized to develop applications that interact with the database.
Low-code platforms often incorporate various drag-and-drop building blocks to help develop applications. These pre-generated components may include tables, textboxes, dropdowns, charts, maps, forms, user directories and other elements. Low-code solutions typically make code under-the-hood accessible and portable, too. Degrees here vary. Some, such as some-code, are more geared toward alleviating operational tasks for skilled engineers.
Low-code solutions also may assist with the surrounding operations necessary to create a fully functional app, such as hosting, deployment and monitoring. Example providers in the low-code spectrum include Appian, Claris Filemaker, Mendix, Outsystems, Quick Base and Salesforce Lightning.
No-Code
No-code is like a driverless car
In an automated car, AI handles all. Driving through traffic and re-routing is automated. You don’t need your hands on the wheel—but you do need a destination. Similarly, no-code is the most usable, non-tech option for application development. No-code is geared toward citizen developers and non-tech users.
No-code takes low-code usability to a new level—100% UI-driven, with pre-built templates and the ability to string together utilities. No-code is popular in connectivity and triggers, taking the burden out of integration. Of course, the more you automate, the more granular control you sacrifice. Many no-code options simply require logins and permissions for business applications you already use. Examples include Zapier, IFTTT, Kissflow, Airtable and Retool.
The trend toward low-code and no-code is industry-agnostic, too, having seen adoption by banking, manufacturing, consumer goods, eCommerce, and other areas, Kharade said.
5 Common Low-Code Scenarios
Now that we understand the variants, what are enterprises building using low-code? Kharade sees four common scenarios:
- Building internal forms: Collecting data in a form, such as for an inspection or audit. May require PDF generation. Non-tech folks with an understanding of their industry could create forms.
- Extending with publicly available forms: Making the form public for registration or lead generation. Requires little added functionality and a no-code user could likely implement.
- Working with databases: Junior engineers will readily understand database mechanics, but it’s not entirely non-tech, said Kharade. For example, skills may be required to synchronize databases with other cloud operations.
- Business process management (BPM): Digitizing the traditional corporate project approval processes. What was once paper-based or email-based is streamlined. It could involve picking up an order, pushing to a manager and a manager approves. Digitizing this process requires a deep understanding of the data structure.
- Purely open applications: Perhaps used for onboarding or internal job postings. Requires a fuller amount of user interaction design, workflows, file hosting and API connectivity. May require skilled developers.
“Building your data structure at the start is extremely critical,” Kharade said. To him, more complex projects will still likely require skilled engineers at some point. But, as you can see above, there are many areas where low-code and no-code could streamline the creation of simple apps and tools.
Low-Code, Demystified
In other pieces, I’ve attributed the rise of low-code to an ongoing developer drought and the emergence of a digital-first economy during COVID. Throughout the pandemic, “there has been a frenzy to digitize any manual process,” noted Kharade. “People are scared even to use a pen someone has used.”
Low-code has in part become a critical option to stand up replacement business applications rather effortlessly, providing citizen developers or non-tech users new digital capabilities. With its surge in popularity, the marketing language around this area is a bit befuddling, and many applications claim to be low-code or codeless.
Of course, you could call any tool with a UI “no-code.” Technically, your iPhone apps are codeless from a user’s point of view. However, what sets these low-code development styles apart is their use in forging internal business applications and automating repetitive software development tasks. By utilizing low-code, traditionally non-tech companies could wrap operational capabilities around basic applications and scale them like tech companies.
Another key to low-code, which I will expand on in a follow-up piece, is solving the integration hurdle with third-party APIs. “It’s difficult for a non-tech user to register an API,” Kharade said. One must understand authentication, OAuth, payloads, HTTP methods, responses and headers just to make a single call.
API-driven services have needed better usability for quite some time. Low-code layers could make consuming a vast array of SaaS more inclusive—a fascinating prospect to problem-solvers of all skillsets.