How to Add Reviews to Unbounce: You Can Test This One

Adding social proof to a landing page is advice everyone gives and almost nobody checks. Unbounce is the one builder in this series where checking is a normal part of the workflow rather than a project, because variants and traffic splitting are the product. So the interesting instruction here is not where to paste the snippet, which takes a minute, but to build the page twice and let the traffic tell you whether the reviews earned their space.

The Custom HTML Element

Drag a Custom HTML element onto the page in the position the reviews should occupy and paste the embed into it. There is no code view to toggle and nothing to escape.

In the Classic builder the element renders as a placeholder box on the canvas rather than executing the code, so the editor will not show you the reviews. Preview and the published page will.

Size and position it like any other Unbounce element. The builder is a positioning canvas, so the element sits where you put it rather than flowing with the content around it.

Build the Page Twice

Make a variant of the page with the review section and one without, then split the traffic between them. This is ordinary Unbounce work and it converts a belief into a number.

Run it until the difference is stable rather than until it is flattering. Landing page tests are noisy at low volume and the first day of data is usually the least reliable day of data.

The likely outcome is a lift, and that is worth knowing precisely rather than approximately. The occasional outcome is no difference, which is also worth knowing because it frees the space for something else.

What Is Worth Testing Beyond Presence

Position is the second test. Proof above the form and proof below it are different pages and they often perform differently, particularly when the form is long.

Format is the third. A compact rating line and a full review wall make the same claim at very different costs in page height, and on a landing page height is expensive.

Do these one at a time. Changing the position and the format together produces a result you cannot attribute, which is the most common way landing page tests waste traffic.

Testing Does Not Change What Is True

A tested page is still a page making a claim, and the claim has to hold. The header rating on the widget comes from the platform aggregate rather than from the cards on screen, so it matches the profile a visitor can open in a second tab.

That property is what makes the proof durable under scrutiny. Optimising the presentation of a number is fine; optimising the number is not, and it is the failure mode that a heavily tested funnel drifts towards.

If a variant wins because it makes the rating look better than it is, it has not won anything worth keeping.

How Do You Build the Widget?

Paste your Google, Facebook or Trustpilot profile URL, preview your real reviews and rating, then choose from over 100 layout styles across the slider, marquee, grid, list, masonry and badge families, and match the colours to the page.

Build two layouts if you intend to test format: a compact badge and a fuller section. Having both ready means the second test starts the day the first one finishes.

Free plan, unlimited widget views, which matters on a page taking paid traffic where a view ceiling would fail at the worst moment.

Where Should the Reviews Go?

Start with the section immediately before the form and treat that as the control rather than as the answer. It is the position most likely to win and it is still a hypothesis.

On a long page, a compact rating near the top as well as the section lower down is a variant worth trying, because many visitors decide before they scroll.

Keep it off any confirmation or thank-you page. There is nothing left to persuade anyone of there.