Block CTO: Code Quality Not Tied to Product Success

The hallowed halls of software engineering have long echoed with the mantra of clean code – elegant, perfectly architected, a thing of beauty in its own right.

It’s a gospel preached in countless textbooks, coding bootcamps, and corporate tech departments.

But what if this sacred cow, this gold standard, is, in fact, entirely beside the point?

What if the meticulous pursuit of code quality is a distraction from what truly matters in building successful products?

This provocative question has been thrust into the spotlight by Dhanji Prasanna, the chief technology officer at Block, the fintech powerhouse behind Square and Cash App.

In a recent appearance on Lenny’s Podcast, Prasanna delivered a statement that could be considered heresy in many engineering circles: “A lot of engineers think that code quality is important to building a successful product,” he observed, “but the two have nothing to do with each other.”

His contention is simple, yet profound: perfect code does not make a great product.

Solving real problems for real people does.

This isn’t just a philosophical musing; it’s a lesson Prasanna claims to have learned firsthand during his tenure at Google, specifically through the acquisition of YouTube in 2006.

He recounts how Google’s engineers, accustomed to their own pristine codebases, were reportedly “horrified” by YouTube’s architecture, deeming it “terrible.” Yet, it was YouTube, not Google’s internally developed Google Video, that soared to become one of the most dominant and successful products in the company’s history.

“It really has very little to do with how well it was architected,” Prasanna asserted, cutting through the conventional wisdom with the sharp edge of empirical evidence.

The true metric of success, he argues, is whether a product genuinely serves its users and addresses a tangible need.

His advice to engineers is refreshingly pragmatic: “Just focus on what we’re trying to build and whom we’re trying to build for. All this code can be thrown away tomorrow.”

This perspective challenges a foundational tenet of software development, where maintainability, scalability, and technical elegance are often seen as prerequisites for long-term success. Prasanna’s stance suggests a radical reorientation, prioritizing user-centric problem-solving above internal engineering aesthetics.

It implies that a functional, albeit imperfect, solution that resonates with users will always trump a technically flawless system that fails to meet a market need.

It’s a bold claim that forces a re-evaluation of where true value lies in the software development lifecycle.

Prasanna’s comments arrive at a fascinating juncture in the tech industry, a period defined by the meteoric rise of artificial intelligence.

On one hand, there’s a chorus of voices, like Google’s head of research Yossi Matias, who insists that “everybody should learn how to code,” viewing basic coding skills as more critical than ever in navigating the AI era.

Bluesky CEO Jay Graber echoes this, arguing that a deep understanding of “good code” is essential for evaluating the output generated by AI.

The implication here is that even if AI writes the code, humans still need the discernment to judge its quality and efficacy, suggesting a continued, albeit perhaps different, emphasis on code literacy.

However, Prasanna’s viewpoint finds unexpected allies in this evolving landscape.

Peter Schwartz, Salesforce’s chief futures officer, has publicly stated that coding is no longer the “must-have skill” it once was. Instead, Schwartz champions empathy and the ability to “work with other people” as the paramount skills of the AI age.

This shift in emphasis suggests a move away from purely technical prowess towards softer skills that facilitate problem identification and collaborative solution-building, aligning neatly with Prasanna’s focus on purpose over perfection.

The specter of AI’s increasing capability to write code further complicates the narrative. Google CEO Sundar Pichai revealed last year that AI is already generating over a quarter of the company’s new code.

This statistic, perhaps more than any philosophical debate, underscores the potential for a paradigm shift.

If AI can handle the technical heavy lifting of code generation, does the human engineer’s role then transform from architect and builder to problem-solver, strategist, and evaluator?

The very definition of “coding” itself may be undergoing a profound transformation.

Prasanna’s message, therefore, is not just about devaluing “clean code”; it’s about re-evaluating the entire engineering mindset.

It’s a call to look beyond the immediate technical artifact and instead focus on the ultimate purpose and impact.

In an era where technological trends can rapidly obsolesce, he reminds us that “technology is here to serve us.” If we have a clear purpose, he suggests, we can make any technology, regardless of its inherent elegance or newness, serve that purpose effectively.

This perspective prioritizes utility and impact above all else, positioning technology as a tool rather than an end in itself.

This isn’t to say that sloppiness is to be celebrated, or that technical debt is a badge of honor.

Rather, Prasanna’s argument is a powerful reminder that engineering, at its core, is a means to an end.

The end, in this case, is a product that genuinely solves a problem for its users.

The quality of the code, in this light, becomes secondary to the quality of the solution.

It’s a perspective that could redefine what it means to be a successful engineer in the age of AI, shifting the focus from the pristine lines of code to the messy, human reality of problem-solving.

It’s a challenging thought for many, but one that demands serious consideration as the tech landscape continues its relentless evolution, forcing a re-evaluation of priorities and skills for the next generation of innovators.