One page or several: the only structural decision that matters

You can spot a stalled personal site from the navigation bar. About, Work, Contact — three links, drawn from the same template everyone starts with. Click Work and there are two projects. Click About and the first paragraph is the homepage intro rewritten slightly worse, because the homepage already used the good version of that sentence. Contact is one email address alone on a screen with a lot of white space around it, doing the job a line in a footer would have done.

Nothing there is broken. Every page loads, the design is fine, and the person spent a weekend on it. What went wrong happened before any content existed: the template offered pages, so pages got filled. The material was one page's worth, distributed across three because the shape came first.

This is the decision worth being deliberate about, and it is almost always made by accident — inherited from a template, a tutorial, or the assumption that more pages read as more serious. They do not. They read as more clicks.

The case for splitting is real, and it is narrower than it looks

The honest argument for multiple pages is not aesthetic, and anyone who dismisses it outright is arguing badly. Three things genuinely require a separate URL.

A page that can be sent on its own. If a client asks for the retail rebrand and you send a link that scrolls them past your bio, your interests and two unrelated projects, you have made the recipient do work you could have done. A case study with its own address is a thing you can put in an email. An anchor on a long page is close, but it lands mid-scroll with no context above it, which is a slightly worse experience for anyone who arrives cold.

A page that answers a different question than the rest of the site. Search engines match pages to queries, and a page about "conversion copywriting for SaaS" competes for something different than a page about you. If part of your work is a service people search for by name, that part earns its own page and its own title tag. This applies to services and named projects. It does not apply to About, which nobody searches for.

And depth that a visitor has chosen. A five-hundred-word account of a project, with process shots and the decision that made it work, is worth writing for the two readers a year who want it. On a single page it is a wall the other ninety-eight readers have to scroll past. Behind a link it costs them nothing.

Notice what all three have in common: they are about items, not about you. Which gives the test.

Count items, not paragraphs

The threshold is whether you have a collection that grows.

A collection is a set of things of the same type where a new one arrives without you planning it: case studies, articles, talks, exhibitions, published papers, open-source projects. The defining property is that a visitor scans it rather than reads it. They are looking for the one entry closest to their situation, and scanning a list of twelve titles is a different act from reading twelve paragraphs.

If you have that, and it is genuinely growing — three now, likely eight next year — you have earned structure, because a growing list on a single page eventually breaks the page. It pushes the contact section below a scroll nobody completes, and the newest work sits wherever the layout happens to put it.

What is not a collection: your biography, your skills, your philosophy of work, your availability. Those are paragraphs. They can be long or short, but there is exactly one of each and none of them will ever become five. Giving a paragraph its own page does not add depth; it adds a click and a heading you now have to write.

The practical version of the test is a question with an uncomfortable answer. If you split the site tomorrow, what appears on page two that could not appear on page one? If the reply is "the same content, with more room", the room was never the constraint.

What the second page actually costs

The cost is never counted, because each item is individually small.

Navigation is the first. One page needs no nav. Two pages need a way to move between them that works on a phone, marks where you are, and looks intentional — a design problem you did not have five minutes ago, and one that most personal sites solve by copying a corporate pattern that fits a company with departments.

Then the redundant intro. Every page needs an opening line, because visitors do not all arrive at the homepage. So the About page re-establishes who you are, the Work page re-establishes what kind of work, and you write your positioning three times in three slightly different ways. The version on the homepage is the one you thought hardest about. The other two are weaker and they are what half your visitors read first.

The third cost is decay, and it is the one that matters in year two. A one-page site has a single place where the truth lives. When your title changes you fix it once, in twenty minutes, and the site is honest again. A five-page site has five places where a fact can quietly go stale, and the failure is not that a visitor spots the contradiction — it is that they cannot tell which parts are current, so they discount all of it. The case for treating a personal site as an address rather than a project rests on exactly this: small sites stay true, large ones rot, and rot is visible.

There is a fourth cost, subtler. A one-page site is finished. It has an end, you reach it, and there is nothing left to fill. A site with pages always has a page that is thinner than the others, and that page is a standing invitation to keep working on the website instead of doing the thing the website is about.

The long scroll with anchors covers most of what people actually want

The middle path is a single page with real sections and named anchors, and it is underrated because it looks less like a website.

Sections do the structural work — a visual break, a heading, a change of rhythm — while the document stays one document. Anchors give you the shareable link when you need one (yoursite.com/#retail-rebrand goes exactly where a page would have gone). A small navigation bar that scrolls to sections rather than loading pages gives the scanner an overview without giving you five files to maintain. And a visitor who simply keeps scrolling gets the whole story in the order you chose, which is the one thing multi-page structure takes away from you permanently.

Where this runs out is depth. If a section wants eight hundred words and four images, it does not want to be a section. That is the moment to split, and it is a much later moment than most people think.

Some tools decide this for you, which is a feature until it isn't

A group of tools makes exactly one page and nothing else, and their constraint maps neatly onto the case above. Carrd is the established one. reach is narrower still: it takes a CV and a photo and composes a finished page from them, live in under two minutes, and one project is exactly one index.html. There is no sub-page, no navigation between pages, no site tree, and no way to add one later. If your site ever needs a second page, you are moving, not expanding. It also has no CMS, so a collection cannot grow inside it; no HTML export and no custom code field, so the markup does not leave with you; and no contact form, only a mail link. Those are real edges and they should decide the choice for anyone who suspects the collection is coming.

What is worth noticing on the price side is how much of what you pay for on a general builder is page capacity you will never use.

Route	Pages	Price	Billing
reach subdomain	one	$0	included
reach Premium (custom domain bought through reach)	one	$4.99	per month, or $49 per year
Carrd Pro Standard (custom domain starts here)	one	$19	per year, annual only
Framer Basic	30	$10	per month, billed annually
Webflow Basic (no CMS at this tier)	300 static	$15	per month, billed annually

Prices checked August 2026. Webflow's CMS starts a tier higher, which is worth knowing if your reason for choosing it was the growing collection in the first place — the cheapest paid plan gives you three hundred pages you have to build by hand and no way to model them.

Nobody with one page's worth of material needs thirty. But if a collection is coming, page capacity is not what you are buying either; a content model is, and that is a different line on the pricing table.

Start at one, and know what the upgrade looks like

Build one page. Get it live — the argument for having something online this week is that a plain page that exists beats a considered one that does not, and the structural decision is easiest to make correctly when you have written the content first and are not guessing at what will fill a nav bar.

Then let the collection announce itself. It will, if it is real: you will have four case studies and the page will feel long, or you will be sending people a link and telling them which part to scroll to. That is the signal, and it is unambiguous when it arrives.

At that point the move is smaller than it sounds, provided the tool allows a second page at all. The existing page becomes the index — it already says who you are and what you do — and each item moves to a URL of its own with the shared framing written once, at the top of the list rather than three times across three pages. Writing is the one collection that should be planned for in advance rather than retrofitted, because a blog changes what the site is and what it costs you every month; that trade is worth settling before you start, and whether the writing belongs on your own site at all is a genuinely open question.

The site that ends up with pages because the material demanded them looks nothing like the site that started with pages because the template had them. The first has a Work section with nine things in it. The second still has two.