In Short
- The relentless chase for the newest, shiniest tech stack is often a developer-centric ego trip, not a client-focused solution.
- True mastery lies in the fundamentals—HTML, CSS, vanilla JavaScript, and core server-side principles—not in memorising the API of this week's hot new framework.
- The 'best' stack is a dangerous myth; the right tool is dictated entirely by the project's actual needs, the client's budget, and their ability to manage the final product.
- For the vast majority of small businesses, proven, 'boring' technology like WordPress and PHP delivers superior value, usability, and long-term viability compared to over-engineered JavaScript monstrosities.
- Stop building for your CV and start building for the end-user. The only metrics that matter are speed, reliability, and a seamless experience—the technology used to achieve them is utterly irrelevant to the customer.
The Cult of Complexity
I see it every week. A new client comes to me, frazzled, holding a digital bag of bolts that used to be their website. It was built by a developer—a smart kid, probably—who was absolutely frothing at the mouth to use the latest tech stack. A headless CMS, statically generated with Next.js, deployed via a labyrinthine CI/CD pipeline, with a GraphQL API for good measure. All for a five-page website for a local bakery.And now? The baker can't update her trading hours, the contact form intermittently fails, the hosting bill looks like a phone number, and the original developer has vanished into the ether to chase an even newer, shinier framework. This isn't progress. It's a self-inflicted wound, driven by a collective industry psychosis I call CV-driven development.We, as developers, have become obsessed with the tools over the craft. We debate the esoteric merits of Svelte's compiler versus React's virtual DOM while the client just wants their phone number to be visible on a mobile phone. Fast. It’s a colossal, navel-gazing distraction.This isn't a Luddite's rant against new technology. Far from it. I've been
building complex online platforms for over two decades. I love a clever solution. But I've learned—often the hard way—that the most clever solution is rarely the most complex one. It's the one that elegantly, robustly, and invisibly solves the user's problem and then gets the hell out of the way.
A Pragmatist's Tour of the Battlefield
So let's cut through the noise. Let's talk about the actual tools on the belt and when—or if—you should ever reach for them, particularly when you're building for the backbone of our economy: small businesses.
WordPress: The Unkillable Workhorse
Ah, WordPress. The platform every 'serious' developer on Hacker News loves to sneer at. It's PHP. It's 'monolithic'. It's for bloggers. They're wrong. Their ignorance is a symptom of the very disease I'm talking about.WordPress powers
over 40% of the entire internet for a reason. It is, without question, one of the most successful
open-source projects in history. Why? Because it solved the client's problem first. It gave normal, non-technical people the power to
manage their own content. That is a monumental achievement.Is it perfect? God, no. A poorly managed WordPress site with 73 random plugins is a security and performance nightmare. But a
professionally built WordPress site—with a custom theme, judicious use of quality plugins, and proper hosting—is a formidable business tool. It's cost-effective, the talent pool to maintain it is vast, and the user experience for the *client* is second to none. For 90% of small business websites—brochure sites, lead generation sites, local e-commerce—starting with anything else is an act of pure hubris.
The JavaScript Hydra: React, Vue, Angular & The New Gods
React is brilliant. So are Vue and Angular. They solved a genuine problem: managing complex state in highly interactive web *applications*. Think Facebook's newsfeed, Google Maps, or a complex SaaS dashboard. For these use cases, they are the right tool. They are magnificent.The problem is, we started using these sledgehammers to crack nuts. We build simple, content-driven websites with them, creating a tangled mess of client-side rendering, hydration errors, and gargantuan JavaScript bundles that make a user's phone feel like it's mining Bitcoin just to display some text. Frameworks like Next.js and Nuxt try to paper over these cracks with server-side rendering and static generation—and they are incredibly clever—but they also add another thick layer of abstraction and complexity. You're not solving the problem; you're just moving it around.Before you reach for a JavaScript framework, ask yourself one question: Is this a document, or is this an application? If it's mostly a collection of pages to be read, you're probably using the wrong tool. You're optimising for your own development experience, not the user's download speed.
jQuery: Don't Call It a Comeback
I can hear the gasps. jQuery? In 2024? Yes. Absolutely. The web development community's collective amnesia about jQuery is baffling. We threw it out because it wasn't 'cool' anymore, and in its place, we pull in a 150kb framework to handle a simple task like toggling a navigation menu—a task jQuery could do in 2kb with a more readable syntax.Modern, vanilla JavaScript can do everything jQuery can, sure. But for quick, simple DOM manipulation on a site that isn't a single-page application, it's often still the fastest, lightest, and most pragmatic choice. It works. It's battle-tested. It's tiny. Get over your ego.
The Backend Brawl: PHP vs. Node vs. The World
The backend is where things get even more tribal. The truth is, for most web tasks, the language you choose matters far less than how you write the code.
PHP: The language that refuses to die because it's just too damn good at its job.
Modern PHP (version 8 and beyond) is a fast, typed, and robust language. It was built for the web, from the ground up. Its shared-nothing architecture makes it incredibly resilient. Paired with a framework like Laravel, it's an absolute joy to work with for building anything from APIs to full-stack applications. The hate for PHP is a tired, outdated meme peddled by people who haven't touched it since 2005.
Node.js: Fantastic for what it was designed for—asynchronous I/O. Real-time chat applications, data streaming, lightweight APIs—it excels. But as a general-purpose web server for content-driven sites? It can be done, but you're fighting the grain. The single-threaded event loop can get blocked, and the dependency management via npm is a house of cards built on a swamp. The `node_modules` folder is the punchline to a joke we all stopped finding funny years ago.
Python, Ruby, .NET: All powerful, all capable. Python with Django or Flask is a solid choice, especially in data-heavy applications. Ruby on Rails popularised conventions that made us all better developers. .NET is a corporate powerhouse, built for enterprise-level applications with deep integration into the Microsoft ecosystem. They all have their place—a place that is, again, usually not a small business website.
What Actually Matters. The Only Things That Matter.
We need to stop asking, "What tech stack should I use?" and start asking, "What is the best possible experience I can deliver to the end-user and the client?"
Speed. This is not negotiable. A user on a 4G connection in a regional town doesn't care about your elegant code. They care that the page loaded in under two seconds. A simple, well-cached WordPress site built on a solid PHP foundation will destroy a bloated, unoptimised React application on this metric, day in and day out.
Reliability. Does it work? Every time? On every browser? Is the contact form going to send the email, or is it going to throw a cryptic JavaScript error because of a dependency conflict? Proven technology is proven for a reason: it has survived years of battle in the wild.
Client Usability. This is the one we always forget. Can the business owner—who is an expert in baking, or plumbing, or accounting, not in Git—log in and change a sentence on their homepage without breaking the entire layout? If the answer is no, you have failed them, no matter how beautiful your code is.
So, What Should a Pragmatic Developer Learn Right Now?
Forget the hype cycle. Build a foundation of granite.
- Master the Trinity. I'm not talking about React, Vue, and Angular. I'm talking about HTML, CSS, and vanilla JavaScript. Master them. Deeply. Understand semantic markup, accessibility, the cascade, the DOM, and the event loop. These skills will never, ever go out of style. They are the bedrock of everything.
- Pick a Workhorse Backend. Learn a server-side language and learn it well. Given its dominance on the web, PHP is arguably the most pragmatic, employable choice. Learn how to interact with a relational database like MySQL or PostgreSQL. Understand SQL. You are not a web developer if you can't manage data.
- Learn a System, Not Just a Library. Instead of just learning React, learn how to build a complete, robust system. That means mastering WordPress development (the proper way—hooks, filters, custom post types, the REST API) or a powerful backend framework like Laravel. This teaches you architecture, not just syntax.
- Learn to Solve Business Problems. The most valuable skill you can have is the ability to listen to a client's needs, translate them into a technical specification, and deliver a solution that provides a return on their investment. The code is the last step in that process, not the first.
Stop being a tool collector and start being a craftsman. The world has enough developers who can copy-paste the latest framework tutorial. What it needs are professionals who can build durable, efficient, and valuable digital assets for the businesses that rely on them. The tech is the means, not the end. And for too long, we've completely forgotten what the end was.
