React
The React problem people bring me is rarely “we need React.” It’s “we built our whole product in React and now it’s slow, heavy, hard to hire for, and nobody can tell us why.” React is a genuinely good tool for the job it’s actually for — complex, stateful, app-like interfaces — and a genuinely expensive mistake when it’s reached for by default. Most of my React work is on one side or the other of that line, and the value I bring is knowing which side you’re on before the money’s spent.
The problems I solve here
The first is the interface that’s earned its complexity and is now buckling. A real application — lots of interactive state, live data, a UI that behaves more like software than a page. It started fine and has become a performance and maintenance problem: re-renders nobody can trace, state scattered across a dozen patterns, a bundle that takes too long to load. The job is to find where the architecture went wrong, usually in how state is managed, and straighten it without a rewrite.
The second, and more common, is the one where React was the wrong call from the start. A content site, a marketing site, a mostly-static product wrapped in a full single-page app for no reason anyone can now defend. It’s slower than the server-rendered version would have been, worse for SEO, and harder to maintain — paying the cost of a complex tool to solve a simple problem. There the honest work is sometimes telling someone the thing they paid a lot for should have been built differently, and mapping the cheapest route back to sanity.
The third is the team that’s drowning in their own front-end stack. State libraries layered on state libraries, a build pipeline only one person understands, every new feature slower than the last. The fix is reduction, not addition.
How I think about React
React earns its place when the interface is genuinely an application: real-time dashboards, complex forms with interdependent state, collaborative tools, anything where the UI is doing serious work in the browser. When the interactivity is real, React’s component model and ecosystem are a fair trade for the weight. I’ll happily reach for it then.
What I won’t do is accept it as the default for everything with a screen. The industry has a habit of building static content as a single-page app, paying for client-side rendering, hydration, and a heavy bundle to display things that wanted to be plain HTML. That’s the most expensive common mistake in front-end work, and I see it constantly. The first question I ask of any React proposal is the unfashionable one: does this actually need to be an app, or is it a document pretending to be one? For interactivity layered onto otherwise server-rendered pages, I’ll usually reach for something lighter like Vue before committing to a full React build.
The other trap is the ecosystem churn. React’s surrounding tooling reinvents itself constantly, and there’s real pressure to chase every new state-management or framework fashion. I don’t. The state-management problem is almost always a design problem in disguise — too much state in the wrong places — and no library fixes a bad data flow. Get the shape of the state right and most of the tooling debate evaporates. Boring, legible React beats a clever stack that needs its author present to maintain.
How I’d actually do it
When a build genuinely needs React, I keep it lean and start from the data, not the components. Most React messes are state messes: the same truth held in three places, drifting out of sync, forcing defensive code everywhere. I’d settle on one clear source of truth, keep server-owned data on the server where it belongs, and let the components stay thin. That alone removes most of the complexity people reach for libraries to manage.
I also build to an MVP. Rather than scaffolding the entire app architecture up front for a scale that may never arrive, I’d ship the one interactive flow that matters most, properly, and grow from there. And the data the interface runs on stays yours — owned in your own backend, served through your own API — so the front end stays a presentation layer over systems you control, not a place your logic and data quietly leak into a structure only the front-end team understands.
Where this connects
This is one of the specialisms I’m hired for. A React front end is almost always a window onto data that lives elsewhere, which is why this pairs naturally with API integrations and with the Laravel or Node backend behind it. When the interface in question is a data view, that’s really a dashboard problem. And when a build’s gone heavy and stalled, the engagement is usually project rescue — getting it back to something maintainable.
When you need a team instead
A large, long-lived React application with continuous feature work is team work, not one-person work. That’s my agency, Digital Royalty — the crew and the process to build and run it at scale, to the same standard. My role is more often the senior judgement on top: the architecture call, the rescue, the direction. I’ll tell you straight which one you need.
Who this isn’t for
If you want someone to add React to a project because it’s the fashionable choice, I’m the wrong call — I’ll probably talk you out of it. I’m worth hiring when you want the honest answer about whether the complexity is justified, and the skill to handle it when it is.
Talk to me
If your React app has gone slow and heavy, or you’re about to build one and want to know if you should, tell me about your problem. Sometimes the most valuable thing I’ll tell you is that you don’t need it.