A/B Testing Links: Compare Landing Pages with a Rotator

You have two landing pages and a colleague who disagrees about which is better. Let your audience decide. A/B testing at the link level is one of the simplest ways to compare destinations: one short link, several possible pages, and click data that shows how each one performs. This guide covers how link-level A/B testing works, how to set it up with ShortGo’s rotator, and how to read the results without fooling yourself.

What link-level A/B testing actually is

Classic A/B testing tools run inside your website and need a script on the page. Link-level testing works one step earlier: when someone clicks your short link, the redirect decides which destination they land on.

That brings a few practical advantages:

• No code on the destination. You can test pages you can’t easily edit, such as two different marketplace listings, two app store pages, or a landing page builder versus your main site.
• Whole-page comparisons. For example, a long sales page against a short one.
• One link everywhere. The same short link can go into an email, a social post, a printed QR code and an ad, and every click joins the same test.

The trade-off is that the redirect only sees the click. Whatever happens after the visitor lands (sign-ups, purchases) has to be measured on the destination, for example with your analytics tool or with UTM parameters that tell the variants apart.

How the ShortGo rotator works

In ShortGo, the A/B rotator is part of a link’s targeting settings. You add several destination URLs and give each one a weight, such as 50/50 for an even split or 70/30 if you want most traffic on the proven page while you try a challenger.

A few details matter when you plan a test:

• Each click is picked randomly by weight. Over many clicks, the split will approach the weights you set. With small numbers, it can look uneven, which is normal randomness.
• It is not sticky per visitor. The same person clicking twice may land on different variants.
• Targeting rules come first. ShortGo evaluates targeting rules (geo, device, language, time) from top to bottom, and the first match wins. Only clicks that match no rule go to the rotator; if there is no rotator, they go to the main destination. If you have a rule sending all mobile users to an app page, mobile clicks will never enter your test.
• Analytics are split by variant. The link’s statistics break clicks down by A/B variant (and by matched rule), so you can compare variants side by side.

How to set up an A/B test in ShortGo

1. Create or open a short link. Pick a memorable custom alias if the link will be shared publicly. Want it on your own brand? See our guide on what a URL shortener is and why branded links matter.
2. Open the targeting settings and add the A/B rotator. Add each destination URL as a variant.
3. Set the weights. Start with an even split unless you have a reason not to.
4. Tag each variant. Add a distinct utm_content value to each destination (for example utm_content=variant-a and utm_content=variant-b) so your website analytics can attribute conversions to the right version. Our UTM parameters guide explains how to keep these consistent.
5. Check your rules. Make sure no targeting rule above the rotator is quietly catching the traffic you want to test.
6. Share the link and wait. Resist checking results every hour.

If you prefer to test whole sets of links, Campaigns offer a related option: each campaign can have a public rotator URL (/r/<slug>) that sends each visitor to a random active link in the campaign.

Designing a test that tells you something

Change one meaningful thing

If variant B has a new headline, price display and layout, a win tells you B is better but not why. When possible, change one major element between variants.

Decide the goal before you start

Clicks on the short link are the same for every variant, because the visitor has not seen the destination yet. Your real metric is what happens after landing: sign-ups, purchases, time on page. Decide which one counts before the test begins, so you don’t pick the metric that flatters your favourite.

Give it enough traffic and time

A common rule of thumb is to run a test for at least one full week, so weekday and weekend behaviour are both included, and to avoid calling a winner from a handful of conversions. ShortGo’s weekday × hour heatmap helps you see whether your traffic has strong daily patterns that a short test could miss.

Keep conditions equal

Don’t change the ad, audience or email copy halfway through. If something external changes, note the date and consider restarting.

Reading the results

In the link’s analytics, filter by A/B variant to confirm the split looks right and to compare traffic quality between variants: countries, devices, referrers. Then move to your destination analytics, where the utm_content tags show conversions per variant.

When you have a winner, you have two options: set its weight to 100 percent, or remove the rotator and make the winner the main destination. The short link stays the same, so every email, post and printed QR code keeps working.

FAQ

Does the A/B rotator remember which variant a visitor saw?

No. Each click picks a variant randomly by weight, so a returning visitor might see a different page. For most link-level tests this has little effect, but keep it in mind when interpreting results.

Can I combine A/B testing with geo or device targeting?

Yes. Targeting rules are checked first and the first matching rule wins. Clicks that match no rule go to the rotator, so you can, for example, send one country to a localised page and A/B test everyone else.

Is A/B testing available on every plan?

Feature availability depends on your plan. Check the pricing page for current details, or create an account to try it.