App landing pages that convert: a checklist

A store page is a good place to install an app, but it's not a good place to persuade someone. You can't change its layout, you can't write a different headline for each ad, and it shows your competitors right under your app. A landing page, a short page between your link and the store, gives you all of that back.

It also adds a step, and every step loses people. This checklist covers when a landing page is worth that step, what to put on it, what to leave off, and how to tell whether it's helping.

When a landing page helps, and when it doesn't

Use one when:

• The ad needs a website. Traffic and website campaigns on Meta, TikTok, Reddit, X or Pinterest need a page to send people to. Sending them straight to a store works, but you lose the chance to repeat the ad's promise.
• Each ad makes a different promise. "Track your lifts in 3 taps" and "Your first 5 km in 8 weeks" deserve different headlines. A landing page can change its headline per link; the store can't.
• It's your link in bio. One page with the install first and your other links (website, Discord, YouTube) below.
• Desktop visitors matter. Someone on a laptop can't install a phone app from a store link. A landing page can show both store buttons and a QR code to scan with their phone.

Above the fold

Most visitors arrive from a video on a phone and decide in a couple of seconds. The first screen, before any scrolling, has to do all the work.

Below that, a few screenshots and up to three short benefits. Then answers to the questions people have before installing: is it free, does it need an account, does it work offline.

Writing the headline

The headline does most of the work, so it's worth a few drafts.

• Say what the person gets, not what the app is. "Know exactly what to lift next session" beats "A smart workout tracker".
• Use their words. Read your store reviews and the comments under your videos. The phrases people use to describe their problem make better headlines than anything written in a meeting.
• Keep it short enough to read at a glance: one line on a phone if you can, two at most.
• Match the ad. If you have three ads with three angles, write three headlines and give each ad's link its own.
• Put numbers in only if they're true and yours: "Free for the first 7 days" is fine if it is.

The subline below it can carry the second most important thing: what makes the app different, or who it's for.

What to leave off

• Navigation. No menu, no links to your blog. The page has one job.
• Cookie banners and trackers. A page with no third-party scripts and no cookies has nothing to ask permission for, and loads faster.
• Long videos that autoplay. If you have a video, show a cover with a play button and let people choose.
• Several calls to action of equal weight. Extra links can live under the store button, smaller.

Speed is a feature

People coming from a social app are impatient, and in-app browsers are often slower than the phone's own. Every hundred kilobytes and every extra request costs some of them.

• Keep the page to one small HTML document if you can.
• Shrink every image to the size it's shown at, in a modern format such as WebP.
• Don't load fonts, scripts or images from other companies' servers.
• Test it on a mid-range Android phone on mobile data, not on your laptop.

Share previews

Your link will be pasted into chats and posts, which show a preview: a title, a description and an image. Set these on purpose (Open Graph tags) so the preview shows your app's promise and a clear screenshot, not a generic store page or a blank card.

Ads and reviewers

Ad platforms review the page your ad sends people to. Show reviewers the same page everyone else sees. Never show reviewers one page and users another (cloaking): it breaks every platform's rules and gets accounts banned.

Measure the page, not just the clicks

A landing page is worth its extra step only if it sends more people to install, and to pay, than going straight to the store would. Measure the steps:

1. Page views: people who arrived.
2. Store taps: people who tapped your store button. Tap rate = store taps ÷ page views.
3. Installs from that page.
4. Paying users from that page.

The first two you can measure on the page itself. The last two need the app to report back, because the store sits in between.

Compare versions, one change at a time

Make two or three versions that differ in one thing: the headline, the first screenshot, the button colour. Point different ads or posts at different versions, and compare their tap rates and, better, their installs and paying users. With small numbers, differences are mostly noise; wait until each version has enough views before declaring a winner.

Doing it with Unolink

You can build all of this yourself, or skip the page builder, hosting and analytics setup. Unolink landing pages are on the Indie plan ($6 a month) and up, and sit in front of the links you choose, so each ad, post or QR code can have its own page.

• Built for this checklist: icon, rating, headline, up to three benefits, screenshots, one store button for the visitor's platform, and both buttons plus a QR code on desktop. Fill in from App Store brings in your icon, rating, screenshots and release notes, so a first page takes minutes.
• A version per link: pick which variant each link shows, so each ad's page can repeat its own promise.
• Blocks: a video shown as a cover that plays only when someone taps it, questions, what's new, and the share preview.
• More links under the store button for your website, Discord or YouTube, so the same page works as your link in bio. For all your apps on one page, there's also the developer page.
• Fast and private: no third-party scripts, fonts or cookies, so no cookie banner, and images sized for phones.
• Up to 3 variants compared: page views, store taps and tap rate per variant, and with the SDK (Pro and up) the installs and paying users that followed. The leader is marked with how sure that is, and it says when it's too close to call.
• Previews done right: chats and posts show your share title and image, and known bots and link previews are never sent to a store and never counted.