Where next for no-code?

Published in

Charge VC

8 min readMay 4, 2020

This post is the first in a series on the no-code movement. Over the next few months, we’ll share our learnings from talking with founders, investors, and makers. To begin with, we explore the question of what makes a company no code.

“The best code is no code at all. Every new line of code you willingly bring into the world is code that has to be debugged, code that has to be read and understood, code that has to be supported. Every time you write a new code, you should do so reluctantly, under duress, because you completely exhausted all your other options.” — Jeff Atwood

The developer dearth

For all the ubiquity of technology in our lives, its architects are few and far between. Though exact numbers are hard to come by given the lack of credentialing required, there are an estimated 30MM software developers in the world, with 3.4MM in the US.

In and of itself, that figure is not particularly remarkable — there are considerably fewer doctors (1.1MM) and lawyers (1.3MM). But when viewed with a sense of demand in mind, the figures are striking. While the US has an oversupply of lawyers and should have a total shortfall of a mere 120K physicians and surgeons by 2032, software developers are in high demand. Before the current slowdown, the US had ~1MM tech openings, with the gap set to widen. On an annual basis, 120K computing jobs should be made available through 2022, while only 40K new graduates will enter the workforce.

This shortage has fundamentally altered the processes of both large and small businesses. As more companies look to build a technological edge in areas like AI, cloud computing, and data science, the fight for engineers has manifested in a willingness to secure talent with elevated salaries, signing bonuses, and other perks. As much as a third of tech openings remain unfilled after 5 months, resulting in significant productivity losses. Intriguingly, this is not a problem isolated to Silicon Valley; salaries grew more quickly outside the tech mecca than within it.

While increasing compensation might be a viable option for the deep-pocketed, it is considerably less so for small and medium businesses. Instead, many are turning to new tools that allow the non-technical to build software.

No-code: a definition

“No-code” has become the accepted umbrella term for the class of visual programming platforms that have risen to prominence over the past few years, allowing “citizen developers” to start shipping software products without writing code. As it stands, the market is valued at $4B, but it’s growing as much as 50% YoY, projected to reach $21B by 2022. Other estimates suggest the market may reach $52B by 2024.

The effect of these platforms can be profound, potentially increasing development speed by 10x, and allowing R&D spend to be allocated more cost-efficiently.

Despite those benefits and the buzz the space has attracted at least in VC circles, it’s surprisingly tricky to answer a simple question: what exactly is no code?

There does not seem to be an entirely reliable definition, or at least not one that stands up to scrutiny. Unquestionably, a no-code tool must remove, or greatly reduce the need to code to solve a problem previously solved through code.

Beyond that, most platforms seem to possess the following traits to achieve that end:

Accessible programming interface
Little-to-no prior knowledge required
Shallow learning curve
Inclination towards action
Limited scope of functionality

Even with the above description, it’s not entirely clear why Webflow is often hailed as a no-code tool but Wix isn’t. In some respect, the no-code movement seems to be a question of expectations, a moving frontier. Once a platform gains mass acceptance, it may no longer have the feel of a no-code product, even if it adheres to the above characteristics.

Just as this makes the act of definition difficult, it also complicates the matter of determining where the most promising applications will be built. While there are many ways to try and answer that question, one approach is to analyze the particular technical skills in-demand. Applications that serve as a substitute for those skills should be especially valuable.

In-demand

While developers are sought after at a general level, there is variation within specializations. Per LinkedIn’s 2019 report, the hard skills most desired include knowledge of cloud computing, artificial intelligence, mobile app development, NLP, scientific computing, game development, software testing, data science, and computer graphics.

A similar report from Indeed unpacks the fastest rising tech skills, as follows:

Across both lists, there is plenty of overlap when normalized. Cloud computing is prominent (Azure, AWS), as is data science (Tableau, Scala, Redshift), app development (Swift), artificial intelligence, and machine learning.

On a salary basis, app makers may be most prized with iOS and Android developers topping average salaries as of January 2020. Other well-compensated skills included specialists in Python, JavaScript, C++, Java, .NET, PHP, and C#. Though the languages and skills above are often general in application — Java can be used to build both mobile and desktop products, for example — they nevertheless overlap with the skills mentioned in the Indeed report.

No-code beginning to fill the gap

With a clear dearth of developer talent, particularly those with certain highly-prized skills, no code is emerging as a viable solution. Platforms that offer a visual programming interface are rushing to fill the gaps, allowing builders without engineering backgrounds to create software to solve business problems.

The Indeed and DAXX reports illustrate the demand for engineers skilled in mobile app development, game development, and data science. Using those categories as examples, we can see how no-code tools are emerging to fill the void.

There are close to 3MM apps on the Android store, with the iOS ecosystem expected to reach 5MM in 2020. Increasingly, we may see makers constructing creations in Adalo, Draftbit, Thunkable, AppSheet, or Glide, rather than slogging it out with Swift, Java, and Objective C, for example.

Game creation appears to be undergoing a similar transformation with tools like Buildbox, Game Salad, Dreamcraft, and Koji allowing makers to construct games of differing degrees of complexity. Google has also gotten in on the act, with “Game Builder” which enables the non-technical to build 3D games. By doing so, these apps remove the need for builders to have a knowledge of C# and C++, among other languages.

Though many will be familiar with incumbent tools like Tableau and Looker in the data science space, Lobe, Gyana and Obviously.ai are less well-known. All offer user-friendly ways for civilians to take advantage of the power of AI and ML without knowing how to write in Python, R, SQL, or Scala.

Opportunities

Because of the amorphous nature of some no-code platforms, it’s difficult to envision precisely what sort of tool will best serve those stepping in to fill the skills gaps. However, there are areas that seem particularly worth tracking over the next 18 months or so.

Engineers are said to spend over 50% of their time on non-coding tasks. While there are plenty of different ways this time-suck might be addressed, solutions that allow devs to streamline the creation of internal tools and automate workflows are of particular interest. These tasks are time-consuming and repetitive but may not be particularly high impact, or represent the best use of limited resources. Already, companies like Zapier, Airtable, Parabola, Retool, and others have come to the fore reducing the need for engineers to spend time architecting databases, APIs, or internal workflows. We expect others to join them offering different feature sets, or focusing on a specific vertical.

Relative to those use cases, both no code testing and data science (AI/ML) feel underpenetrated. While companies like Reflect and Testcraft are providing tooling in testing, and as discussed, Lobe and Obviously operate data science platforms, few other solutions abound. This may be because data science workflows have traditionally been ad hoc, relying on custom code. As the importance of data science and AI increases, accessible interfaces that allow for replicability and collaboration will become ever more relevant.

Creativity will abound. Products like Notion defy easy categorization, as does Clay, Coda, and others that fundamentally change the way we work. We hope (and expect) others to play with the boundaries, combining disparate functionality and innovating on product.

Whatever the tool of choice, whatever product gains popularity, we are witnessing a fundamental shift in which the circle of those that build the future is widening. Software is about to have many new architects.

— -

If you’re building something in this space, we’d love to talk. team at charge dot vc.

— -

Companies listed (L →R within categories)

Web

Mobile

Agencies

Testing

Automation

Spreadsheets

Voice apps

Super docs

Community

Data science

AR/3D

Monetization

Gaming