Next.js Development for Content-Led Platforms
When the front end has to outlive the launch.
Next.js is the default answer for a React front end, which is exactly why the interesting decisions are not about the framework. They are about what sits behind it, how much of the page a person who is not a developer can change, and whether the thing is still fast in year two.
Two Next.js front ends in production — the LA28 venue hub for Visit California, and Sound Dollar’s editorial platform, which went from nothing to launched in two months.
Next.js is the default React framework — server and client rendering in one model, image and font optimization, static and incremental generation. Choosing it is close to a non-decision now, and the risk that follows is treating the rest of the stack as equally settled. A front end is only as good as the content model it renders and the editorial workflow behind that — get those wrong and you have shipped a fast site no one can update, which is a slower site within a year. The work that matters is the seam. Mapping structured content onto components so an editor’s change lands where they expect it, keeping the data layer honest about what it fetches, and keeping performance and accessibility measured continuously rather than at launch, because both decay quietly and neither shows up in a ticket.
Our Work with Next.js
How the Work Splits
A React framework with server and client rendering in one model, file-based routing, image and font optimization, incremental and static generation, and a deployment story that most hosts now support natively.
The decisions the framework leaves open — content modelling that maps cleanly onto components, a data layer that fetches what the page needs and no more, editorial workflows that let a non-developer change what they expect to be able to change, and performance and accessibility as ongoing measurements rather than launch-day numbers.
Front ends that survive their second year. On the LA28 venue hub for Visit California, Next.js renders faceted venue search over structured content with automated weekly runs where accessibility and best-practices scores have held at 100 on every recorded run. On Sound Dollar, it took a content business from nothing to a launched editorial platform in two months, scoped deliberately to what had to exist before the writers could start.
The work in practice
No one needs persuading to use Next.js any more, which is the problem. When the framework choice stops being a decision, teams stop making the decisions around it — and those are the ones that determine whether the site is still good in eighteen months.
What the framework does not decide for you
How much of the page a non-developer can change. This is a content-modelling question before it is a front-end one. Components that map cleanly onto structured content give an editor real control; components assembled from a dozen loosely related fields give them a support ticket. The model and the component library have to be designed against each other, not in sequence.
What the page actually fetches. Server rendering makes it easy to pull more than the view needs and never notice, because the cost lands on a server rather than a phone. It still lands. The data layer deserves the same scrutiny the bundle gets.
Whether it is still fast next year. Launch performance is the easy part. Performance is a property that decays — a new embed, a heavier hero, a third-party tag no one remembers approving. Unless something measures it on a schedule and someone reads the result, it degrades invisibly.
And whether it is still accessible. The same decay, with higher stakes and less visibility. A component library that met the standard at launch drifts as it grows, and the drift is only ever caught by a check that runs without being asked.
Two front ends, two different problems
The LA28 venue hub, for Visit California. Next.js over structured venue content, with faceted search and an interactive map — national delegations scouting California for training and meeting venues before the 2028 Games, often browsing from another continent or standing on a venue floor. A seven-week MVP proved the shape and the platform grew from there. Automated performance runs go out weekly across every template, and accessibility and best-practices scores have held at 100 on every recorded run since monitoring began. That is the number worth quoting, because it is the one that shows the thing did not rot.
Sound Dollar. A content business with financial planners, journalists, analysts and researchers already assembled and already being paid, and nothing for them to publish into. Every week before launch was expertise bought and not published, and personal finance is a calendar business — miss tax season and you wait a year. Next.js over a headless CMS, scoped hard to what had to exist before the editorial team could start work rather than to the platform the business intended to become. Live in two months.
Both are written up — the venue hub and Sound Dollar.
If you are starting one
The framework will not be your problem. Ask instead who is going to change this page in eight months and whether the model lets them, what the page fetches that no one looks at, and what runs on a schedule to tell you it is still fast and still accessible. A Next.js build that answers those three is a good one almost regardless of how it is written; a build that answers none of them is a rewrite with a date on it.
Practice leadership
Jack Hartman
Jack Hartman has run delivery on complex platform engagements, overseeing the people and processes behind them. He has led front-end decisions judged on the years after launch.