SaaS Productivity in AI World

Artificial intelligence is profoundly altering software engineering within SaaS organisations, shifting the baseline from manual coding to automated task execution and architectural orchestration. While AI assistants and agents yield average individual task productivity gains of 15% to 30%, their holistic impact on SaaS engineering velocity introduces a "productivity paradox" regarding software quality and maintenance.

The structural impact of AI on SaaS product engineering productivity unfolds across several core dimensions:

Acceleration to Development Lifecycle (AI-DLC)

  • Rapid Prototyping and Sprints: Engineering teams use agentic systems to translate high-level business ideas into technical requirements, automatically generating the initial code drafts, test suites, and product documentation. This drastically reduces customer cycle time—the speed at which a feature moves from a request to active customer deployment.

  • Shift to System Orchestration: Engineers spend fewer hours typing line-by-line code. Instead, time is spent managing autonomous agent workflows, defining operational constraints, and designing scalable system architectures.

  • Debugging and Refactoring Gains: Real-world data reveals that the single greatest time-saver for developers is not code generation, but using AI for stack trace analysis, complex error interpretation, and legacy code refactoring.

The AI Productivity Assumption

This code generation acceleration has given the impression that overall engineering outcome has greatly improved. The truth is far from the headline news. A simple internet search or in today’s world, ask AI this question - “How is AI impacting SaaS product software engineering productivity?”. And you get a good reasonable answer as to how AI can enable a more productive software engineering team. However, just like any AI output, the hallucination or the lack of context can provide an answer with some blindspots.

The context here for Software Engineering (SaaS) Productivity metric should be viewed in accordance to the full software delivery lifecycle. A simple model is shown below in this diagram.

The AI Productivity Paradox.

 

Most of AI provided answers to SaaS Productivity concentrates on the latter part of the SDLC, namely the Build-CI-CD loop. This is just a subset of the real delivery flow that encompasses far more steps from the start of the Ideation all the way to a finished Product.

So while I don’t disagree with suggestions on how AI can enable better productivity in the Build-CI-CD loop, the bigger Productivity “Gain” is to be found to the left of the lifecycle.

Most SaaS organisations will measure Productivity based on:

  • How fast you go from the Ideation phase to an actual finished Product (released into Production). Usually this means the key measurement here is Speed, via Velocity metric. Or

  • How efficient you go from the Ideation phase to an actual finished Product (released into Production). This can be the Kanban board where Number of Stories Done is a key measure.

“By focussing AI enablement on the Build-CI-CD loop, can we safely assume, most SaaS organisations misses the true Productivity uplift, initially?”

Don’t Forget Key Questions

 

  • It all begins with an idea. AI enabled delivery can launch a new product or features much quicker nowadays. Perhaps a wiser thing to do, is to step back and ask “What is the value of this work?”. Whatever it is, the way business prioritise higher value work first, will make all the difference.

  • Don’t build it first. Once value is locked in, and priority assigned, prototype it! Pilot or show it to a select group of customers. Customer feedback, and continuation of product/idea iteration is highly valuable. Creativity and innovation processes truly kick goals. Outcome is then measured and reviewed together with all others ideas.

  • Once the customer feedback, needs and business values are all understood, that gives the team a strong confidence to finish building a scalable and reliable product. The difference now instead of weeks or months, the whole process from Idea-to-Prototype-to-Product is just in days. And it hits your organisation goals.

    Therefore the Product Funnel is so important. It is not just about building more Products or Features. Teams may look more efficient and productive, but if it does not move the business goals or lift the revenue, it is all work no gains. It is even more important in today’s world of faster AI Enabled SDLC that choosing the right idea or feature to build, will achieve the right outcomes.

 
The P Formula