Investments in AI are paying off, as organizations across industries report tangible results. The latest EY AI Pulse Survey finds that 97% of senior leaders are seeing positive return on investment (ROI) from AI, including gains in employee productivity (82%), product innovation (80%) and operational efficiencies (79%).
Nowhere, perhaps, does AI offer more potential than an organization’s software development function, with AI-based coding assistants and other tools improving efficiency and reducing cycle times. While these gains are impressive, AI’s real impact on software development will be achieved as organizations shift from a traditional project-focused approach to a product-oriented development model.
This transition requires a lot more than simply bolting AI onto existing software development processes and workflows. The journey from the traditional software development lifecycle (SDLC) to an AI-native product development lifecycle (PDLC) requires a reimagining of processes, supported by important changes to existing culture, operating models and skills.
For many organizations, the rewards will be well worth the effort. Teams who adopt PDLC — an AI-orchestrated, intent-driven model for delivering complete software products, including code, tests, documentation and infrastructure, in days instead of months — are poised to capture a variety of business benefits. Those gains include dramatic improvements in efficiency, the ability to reshape and uplevel traditional software engineering roles, and the ability to create better products that reach customers faster than ever before.
“The most important part of PDLC is that it allows you to build better software,” says Rakesh Malhotra, Principal, Digital & Emerging Technologies with Ernst & Young LLP. “You can iterate and incorporate feedback much more quickly, which makes it more likely that the software you build is compelling to the users you’re building it for.
“PDLC is about building the entire apparatus required to deliver great software — not just the code, but the documentation, the sample code, the tutorials, the marketing collateral and the training that go along with it,” Malhotra explains. “Theimpact of that is an operating model that is orders of magnitude more compelling, efficient and valuable than what we have today.”
How a PDLC transformation goes beyond faster time-to-market
The impact of moving from a project- to product-focused operating model is felt well beyond project timelines. In some cases, PDLC changes decision-making hierarchies. Traditional silos between engineering, design, manufacturing and marketing become more fluid, creating a collaborative “mesh” vs. a sequential workflow.
“As a product manager, engineer, designer or domain expert, you may sit in different organizational structures, but you have to work across functional boundaries effectively to be successful,” says Malhotra.
Skills will also need to evolve, as AI takes over traditional code-writing tasks.
“Programming is a solved problem with AI,” says Malhotra. “The smartest engineers in the world know they don’t need to write code anymore.” Instead, he says, engineers will take on more of a product manager type of role, focused on ideation, intent and outcomes.
“That’s clearly where the value lies,” Malhotra adds. “We’re seeing a new era of full-stack engineers who operate with a lot of cross-functional versatility.”
Consider these three steps to begin a PDLC transformation
Unlike traditional project-led software development, which involves lengthy planning cycles and discrete budget, a PDLC framework lends itself to continuously iterating, evolving and nurturing products and services. And despite some of the broader cultural and organizational challenges, Malhotra maintains it’s relatively easy to get started on the PDLC journey. Consider these three steps:
- Start with a small team and a real-world problem that needs solving. Use PDLC methodologies to quickly create a prototype and then iterate until you achieve a tangible business outcome that gets people’s attention. “Don’t invent a fictitious challenge to point the PDLC at,” says Malhotra. “Use something that can really move the needle and build credibility with the rest of the organization.”
- With a successful proof of concept in hand, you can then build a center of excellence to scale the operating model. Here, teams will define key performance indicators and expected results. “The conversation generally starts with cost savings and efficiency, where you can see dramatic improvements,” says Malhotra. “From there, you can move on to define what better software looks like.”
- Define the guardrails for the new operating model. The ability to build virtually any software product quickly is a significant benefit, but it also carries outsized risk. So, it’s important to be judicious about what you build and how you build it. In addition, risk controls and other governance policies should be embedded in PDLC workflows — including the large language models that fuel the programming process — to help ensure compliance. “There’s a big opportunity to fully automate governance as part of the SDLC process, as opposed to something you add after you build,” says Malhotra.
How PDLC can change your organization’s culture
AI-native software development is fundamentally reshaping how organizations design, build, and deliver products. The immediate benefit is speed: Reducing product cycles provides an obvious go-to-market advantage against competitors still using traditional SDLC methods.
But moving from SDLC to a PDLC model is more than just a process improvement. It’s also a catalyst for new ways of thinking and working, helping to create a product culture focused less on deploying code and more on delivering compelling experiences to both internal and external customers.
“PDLC allows you to build precision software that’s customized for every business unit, for every team, and caters to how users actually work,” says Malhotra.
It also opens the door to the art of the possible. By combining advanced AI with proven product methodologies, PDLC can help transform traditional processes, reshape roles and create a more productive environment for innovation.
“If you could build anything instantly tomorrow, what would you build?” Malhotra asks. “It’s actually a hard question for leadership teams to answer. But that’s where you create durable competitive advantage.”
EY.ai PDLC is a catalyst for new ways of thinking and working, helping to create a product culture focused more on aligning business and delivery than on syncing moving parts. Organizations interested in transformational change can leverage a tailor-fit collection of packaged offerings to adopt and scale the PDLC framework.
Learn more about how EY Consulting can help CIOs rewrite the delivery playbook with an AI-orchestrated product development lifecycle.
The views reflected in this article are the views of the author and do not necessarily reflect the views of Ernst & Young LLP or other members of the global EY organization.