Back in 2011, entrepreneur and investor Marc Andreessen famously said, “Software is eating the world.” He meant that software companies and their solutions are taking over increasingly larger parts of our global economy. He was right. But it is taking a long time because creating software is so slow.
There is little doubt of the importance of development productivity. Strategy-consulting firm McKinsey & Co. has observed that “superstar firms” spend two to three times more on R&D relative to their industry peers. Additionally, 54 percent believe their technology organizations’ performance will significantly or completely differentiate their business from their competitors.
What this boils down to is that literally every Global 2000 firm is now continually involved in the processes of digitization and digital transformation. Naturally, all of this takes software—lots of it. The entire world, it seems, is engaged in a titanic push to build more software. Businesses desperately want to drive increased productivity from the significant dollars they spend on technology development.
In a nutshell, everybody wants more, better software faster.
Yet even today, building software remains something of an artisan craft. It’s a highly skilled trade that has traditionally involved making things “by hand”—in the case of software development, that means line-by-line coding.
As a venture investor, my exposure to innovation companies reveals competitive battles in categories focused on how to increase software development productivity. Following are several tools, practices, and methodologies that are becoming game-changers in terms of how software is being designed and built.
Low-code/no-code development platforms. These tools are exactly what they sound like—they enable software to be built with minimal to no coding, dramatically compressing development time. The basic idea is to build applications using pre-developed components as opposed to building everything from scratch using programming languages. Low-code and no-code platforms have given rise to “citizen developers,” non-programming business users who are building their own software to suit specific needs.
For example, consider the business analyst replacing traditional spreadsheets with custom-built reports to boost departmental productivity. It’s a way to get more customized software than commercial off-the-shelf solutions offer and also eliminates the bottleneck of internal software development requests by enabling business users to effectively “build their own.”
Low-code and no-code platforms have been around a while but they’re still a growing software category. Gartner recently commented in a meeting with BIP Capital that it tracks 320 companies in this space. Some of the most popular platforms are those from Mendix and OutSystems.
Better software-testing practices and tools. Traditionally, the software-development lifecycle was a linear or “waterfall” process that typically went from requirements, design, and implementation to verification and maintenance. This meant that software testing occurred only near the end of the development lifecycle. If bugs were found, developers had to go back in the process, extending production time. Agile methods shifted this process into small sprints, but the testing still typically occurred at the end of a sprint or after integration.
This dilemma has given rise to “shift-left testing” in which software is tested often and in every phase of the development process from beginning to end. With this type of testing, bugs and errors found are usually smaller and are less costly and time-consuming to fix. BIP Capital’s own investment in Cypress.io is an example of shift-left, end-to-end testing technology.
Another practice is “shift-right testing” that focuses on the software’s performance in real-world environments. While shift-right testing occurs near the end of the development lifecycle, it is tied closely to DevOps, a combination of practices and tools that allow organizations to develop better products, faster. Shift-right testing also involves other stakeholders from the beginning, such as product managers and business users—people who are closer to the problem the software is intended to solve and understand what it should do (those “citizen developers”). This again is leading to the democratization of software creation by empowering less technical users to drive innovation.
Use of AI and machine learning. Finally, as the industry continues to evolve, I see and anticipate more use of AI/ML in software development and testing tools in order to speed the product development lifecycle and enhance software quality and capabilities. Just as AI/ML are transforming nearly all business practices, they have the potential to fundamentally transform how software is created, tested, and continually improved upon. BIP Capital portfolio company Kobiton is an example of early innovators in the AI/ML software development space for mobile software testing.
Across all these areas, there are new technologies appearing and significant capital investments being made in the never-ending drive for greater software development productivity. I look forward to seeing what innovations will rise to the top to better enable tomorrow’s software creators.