Skip to main content

Launch a coming-soon page while the rest of your site stays private

You have a domain, a half-built site, and a launch date. You want visitors to land on something deliberate — "here's what's coming, tell me when it's ready" — while the forty pages behind it stay out of Google until you're happy with them.

This walkthrough is the whole loop: build the page → make it home → hide the rest → collect signups → flip everything back on launch day. It's the same pattern Aglyn uses on its own site.

Plan availability

Everything here works on Free: screens, forms, the inbox, per-screen visibility, and the site-wide search switch. Writing signups into a dataset (step 4) needs the data store, which unlocks on Starter and above — the inbox copy is always kept either way.

1. Build the coming-soon page​

Create the screen first; you'll point your domain at it in step 2.

  1. Go to Screens → New screen. Title it Coming soon. Give it any slug for now — coming-soon is fine — you'll change it to / in the next step.
  2. Open it in the Besigner and put the essentials on it. A coming-soon page that only says "coming soon" wastes the visit. Aim for:
    • What's coming, and roughly when. One sentence and a month beats a countdown timer you'll have to move.
    • A notify-me form — see below.
    • How to reach a human. An email address or a contact link; people who arrive early are often the ones who most want to talk to you.
    • Links out — your documentation, your changelog, your social accounts, wherever you already post. These give an early visitor somewhere to go instead of bouncing.
  3. Publish the screen.

The notify-me form​

Add a Form from the element picker with a single Form Field of type Email, and set the Form's Success message to something that confirms the signup ("Thanks — we'll email you when we open"). The Forms & lead capture page covers the form elements and every field property in full; the only thing that matters here is that you give the field a Field name you'll recognise later, such as email.

2. Make it the home page​

A screen's Slug decides the path it's served at, and the slug / means the home page. So swapping your home page is two slug edits.

Change the slug, don't delete the screen

Move your existing home page aside rather than deleting it. Its design, its content, and its version history are all worth keeping, and you'll want it back in step 5.

  1. Open your current home screen (Sites → your site → Screens → click the row). In the Publishing card, change Slug from / to something you'll recognise — home works — and choose Update route. The screen stays published, just at /home now.
  2. Open the Coming soon screen. Change its Slug to / and choose Update route (or Publish, if you haven't published it yet).
  3. Load your site. The coming-soon page is now what visitors get.

A screen can be the home page or have a parent, but not both — if the slug won't take, check that the screen sits at the top level of your screens list.

Changing a slug does not create a redirect

Aglyn rewrites its own routing map when you change a slug, so links within your site keep working. It does not leave a redirect behind at the old path — that URL simply stops resolving. If the old path was public and people have linked to it, create the rule yourself with the redirect manager (a paid feature) — see Create a redirect.

Your unfinished pages are still published and still reachable. Two controls decide whether search engines index them, and which one you want depends on how far along you are.

While nothing is ready: the site-wide switch​

Go to Setup → SEO and turn on Discourage search engines from indexing this site. One switch, whole site:

  • robots.txt refuses every crawler (Disallow: /) and stops naming your sitemap
  • sitemap.xml is served empty
  • every page — including your coming-soon page — carries <meta name="robots" content="noindex">

This is the right choice while the site is a building site. It also covers pages you haven't created yet, which per-screen visibility can't: a screen you publish next Tuesday is indexable by default, so a site hidden page-by-page slowly un-hides itself as it grows.

While the switch is on, a warning banner follows you through the console on every page of that site. That's deliberate, and it's the answer to the most common support question this feature creates — see step 5.

Once you're launching page by page: per-screen visibility​

When the site is mostly ready and you're releasing sections, turn the site-wide switch off and hide the stragglers individually. Open a screen (Screens → pick it) and use the Visibility dropdown in its Page Access card:

VisibilityReachable byIn search
Publicanyoneyes
Unlistedanyone with the linkno
Password protectedanyone with the passwordno
Members onlysigned-in membersno

Unlisted is the one you want for a finished-but-not-announced page: it stays live so you can share the link with a reviewer, and it's kept out of both search results and your sitemap.

noindex is a request, not a lock

Unlisted and the site-wide switch ask search engines not to list a page. Well-behaved crawlers honour that; nothing stops a person with the URL from opening it, and nothing stops a badly-behaved crawler either. If a page must be genuinely inaccessible, protect it properly — see Site protection and Password-protect a screen.

4. Collect the signups​

Every form submission is filed in your Inbox under the form's name, on every plan. That alone is enough to email people on launch day.

To get a list you can actually query, point the form at a dataset:

  1. Create a dataset on the organization Data page → Add dataset, called Launch signups, with a field for the address — email.
  2. Back in the Besigner, select the Form container and set Write to dataset to Launch signups in the Attributes panel.
  3. Select the email Form Field and set Maps to schema field to email.

That mapping is what makes the record land, and it's worth doing explicitly. The picker stores the dataset field's stable Reference ID, so renaming the field later never breaks it.

Left on None (match by field name), the value falls back to matching a dataset field whose Reference ID — the snake_case key beside each field in the Schema dialog, not its Display name — is identical to the form field's Field name. So a form field named email lands in a dataset field whose Reference ID is email with no mapping at all, while one named Your email silently doesn't.

Values that match nothing are dropped from the record without failing the submission, and the same is true of a deleted dataset or an exceeded record quota — the append is best-effort by design, so a data-store problem never costs you a signup. The inbox copy always keeps everything, which is why it's the thing to check if a record looks wrong.

The survey walkthrough covers datasets, schema types and the form pickers in much more depth.

5. Launch day: reverse every step​

Turning a site dark is easy to do and easy to forget. Here is the whole reversal — work down it, then verify with step 6.

  1. Allow search engines again. Setup → SEO → turn Discourage search engines from indexing this site off. The console banner disappears, and robots.txt, the sitemap and the page meta all return within about a minute.
  2. Set the hidden screens back to Public. The site-wide switch and per-screen visibility are independent — turning the switch off does not promote an Unlisted screen. Go through Screens and set each one's Visibility back to Public in its Page Access card. Anything you leave Unlisted stays out of search, silently and indefinitely.
  3. Swap the home page back. Open Coming soon, change its Slug to something else (coming-soon) and Update route — or Unpublish it if you're done with it. Then open your real home screen and set its Slug back to /.
  4. Verify — step 6. Don't skip it. Every step above is a setting that can be half applied, and the failure is silent.
"Why is my site not on Google?"

Nine times out of ten it's this feature, left on. Check, in order: the Setup → SEO switch, then each screen's Visibility. Then remember that re-indexing is the search engine's schedule, not yours — a site that was noindex for a month does not come back the afternoon you fix it. Allow days, and use Google Search Console to request indexing if you're in a hurry.

6. Verify it actually worked​

Three things a crawler reads, and how to read them yourself. Run these against your real domain.

robots.txt — visit https://your-site/robots.txt, or:

curl -s https://your-site/robots.txt

Hidden looks like this — note that no sitemap is named:

User-agent: *
Disallow: /

Visible looks like this:

User-agent: *
Allow: /
Sitemap: https://your-site/sitemap.xml

The page's robots meta — open the page, View source, and search for robots. Or:

curl -s https://your-site/ | grep -i 'name="robots"'

A hidden page returns <meta name="robots" content="noindex, follow"/>. A visible page returns nothing at all — no robots meta means "index me", which is the default.

The sitemap — visit https://your-site/sitemap.xml, or:

curl -s https://your-site/sitemap.xml | grep -c '<loc>'

That counts the URLs you're submitting to search engines. It should be 0 while the site-wide switch is on, and after launch it should match your public pages. Any page you expect to see and don't is a screen still set to Unlisted.

Give it a few minutes

These three files are cached for about five minutes. If a change hasn't shown up, wait a beat and re-run rather than changing another setting.