Shivani Govil, Product Leader at Google.
Generative AI (GenAI) has sparked a market explosion. The application of GenAI to software development is ranked as one of the top use cases of the technology, with the market size expected to reach nearly $3 billion by 2030, growing at a CAGR of 35.3%.
GenAI Uniquely Equipped To Help Software Development
There are a few factors leading to this. The first one is because of the scale of the problem. Almost every company today is a technology company that relies on technology as a fundamental part of the business and employs a number of developers and engineers in their workforce.
Second, there is a finite capacity of people with the requisite skills needed to do the job. In 2022, there were an estimated 26 million professional software engineers globally, and Korn Ferry research forecasts that by 2030 the labor-skills shortage for technical talent will reach 4.3 million workers. This has an impact on a company’s top line and the same report estimates the unrealized output due to this shortage is estimated to be nearly $450 billion by 2030. The good news: GenAI technologies can help fill the gap in this anticipated shortage, both by making the current set of developers more productive and by closing the gap in the number of developers needed in the future.
Third, developers tend to be high-cost resources, and saving a developer even an hour a week through the application of GenAI results in productivity savings that can quickly add up and have a serious impact on your bottom line.
Tackling Core Development Steps
While the software development process is complex, it is a means to an end—the end goal being the ability for businesses to deliver digital experiences that generate business value. The software development journey breaks down into a few major steps: coding, testing and debugging, collectively referred to as the inner loop, and releasing, deployment and management, referred to as the outer loop.
Historically, high-performing development teams have been optimizing each step, creating an efficient frontier for productivity and velocity. GenAI is disrupting this efficient frontier by creating new ways to not just improve each step but also reconnect these steps in more effective ways for more transformative outcomes.
Much of the investment today is directed toward how to assist software developers with tasks they need to perform on a daily basis—like code generation, code completion, transform code, debugging, monitoring and bug fixing. Applying AI to coding tools, which is one of the most popular areas of application, is just the start and only scratches the surface. GenAI in coding tools is expected to represent close to $30 billion by 2032, which is only about 12% of the expected market share for GenAI software development. There is a huge, and largely untapped, opportunity with GenAI in the outer loop, although we are starting to see applications emerge for some outer loop tasks, particularly debugging, monitoring and risk management.
As we start to transform each step of the software development process, there is also the opportunity to rethink the entire product development process. Agentic approaches, which focus on creating agents capable of performing tasks independently, can help with breaking down each step and putting together the right building blocks to piece the final solution together.
Here are a few key steps leaders can take to make sure they are on the right track with the GenAI development:
1. Evaluate the areas where AI can be applied across their SDLC (e.g., coding, testing, etc.).
2. Begin with some pilots/ test efforts to ensure you have the right data, models and implementation.
3. Conduct user research to understand the outcomes of the pilot.
4. Determine how to scale (both from technical and people aspects).
5. Think through the business considerations (ROI, ethics, safety, etc.).
Imagine if software code could be developed based on designs or based on product requirements. In those cases, product managers (who listen to and assess customer needs) could create code based on the requirements of the users—enabling a much larger number of players to be involved in the software creation process and a much faster time to market.
These types of scenarios empower businesses to iterate solutions faster, test product-market fit quicker and launch compatible products to market quicker. They also enable companies to seize market opportunities and release features faster than they do now, fueling growth and profitability from both top-line revenue and bottom-line savings—a truly game-changing equation in today’s hyper-competitive world.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
