What makes a landing page testimonial believable. A bird that remembers faces

· Founder of Magpie AI

Published · 6 min read

In short

A testimonial's believable when a real person a visitor could look up says something specific. Give each quote the person's name, their role or company, a real photo and a link to where they said it, and keep its words to a concrete use or result. Vague praise from nobody in particular is easy to fake, visitors discount it, and invented quotes are banned in the US and UK.

Contents

My rule's blunt. If no real person stands behind a testimonial on your site, I'd take it down, because a quote nobody can trace does more harm than the empty space it leaves.

I know that sounds harsh. Here's why I think it, and what I'd put there instead.

What visitors already assume about your quotes#

They start out doubting them. In Nielsen Norman Group's 2016 study of trust, people treated the quotes and case studies on a company's own site with caution: they wondered whether they were real, and they assumed a company would only ever show the nice ones. The researchers put it bluntly, saying "people trust testimonials from external sites more than those listed on the website itself."

So a testimonial doesn't start at neutral. It starts a little behind, and every detail you give it either earns that back or proves the visitor right.

Why vague praise reads as fake#

Because that's what fakes sound like, and people can't tell the two apart by reading them. In 2011, Myle Ott and colleagues at Cornell put 400 real hotel reviews beside 400 invented ones, written for a dollar each on Amazon Mechanical Turk by people told to pose as happy guests.

Three human judges tried to sort 160 of them. They scored between 53% and 62%, which the authors called "roughly at-chance".

Detail is what gave the game away. The real reviews talked about the place itself (the size of the room, the bathroom, where it was), while the made-up ones drifted off to things outside the hotel, a husband or a business trip, and preened themselves on superlatives. That was hotels rather than software, and I wouldn't push the comparison too far. Still, you'll recognise the pattern on plenty of landing pages.

I see the same shape a lot in our audits. A row of quotes says "Love it", "Best tool ever" or "Five stars, would recommend", signed with a first name or no name at all. None of them has to be fake. A visitor just has no way of knowing, and the words don't give them anything to go on either way.

The four things that make a testimonial believable#

A visitor should be able to check who said it, and learn something they couldn't have guessed. That comes down to four things.

  1. A full name. A first name alone beats nothing, and some customers won't give more than that. A full name is something a visitor can actually look up.
  2. A role or company. It tells the visitor whether this person's like them. "Head of growth at a 12-person agency" places the quote, and "Happy customer" doesn't.
  3. A real photo. A face makes it feel like a person said it.
  4. A link to where they said it. Their own post, a public profile or a review on an outside site. This one deals with the doubt from the section above head on, since the visitor no longer has to take your word for anything.

After that, the words need one concrete thing in them. What they used it for, what changed, or a number they'd stand behind.

Here's a made-up plant care app at fernly.example, before and after:

BeforeAfter
"Love it, best app ever." Sarah"I used to lose a fern every winter. Two winters in, all 11 of my plants are still alive." Sarah Okafor, nurse, Leeds, with her photo and a link to her own post
"Fernly is amazing for our team." Mark"We look after the plants in three offices. Fernly cut our watering rounds from daily to twice a week." Mark Ellis, office manager at Brightdesk, with his photo

The after versions gush less. They're also far easier to believe, and I'd make that trade every single time, even though it means going back to customers and asking them for a sentence they have to think about.

The photo is the part to be careful with#

In a 2012 set of experiments, Eryn Newman and colleagues showed people claims about unfamiliar names, some with a photo of the person and some without. The photos proved nothing either way. People still judged the claims true more often when there was one.

That cuts both ways. A real customer's photo helps their quote land, but a stock photo on a made-up quote borrows trust it hasn't earned, and I doubt a sceptical visitor who's seen that face somewhere else trusts anything near it again. If you haven't got their photo, use the name, role and link, and leave the picture out. An empty circle with their initials is fine.

Where the law draws the line#

Making up a testimonial is now against the law in both the US and the UK. The FTC's rule on reviews and testimonials, in force since 21 October 2024, bars a business from writing or selling one that misrepresents that the person exists, that they used the product, or what their experience was. It also covers quotes a business spreads when it "knew or should have known" they were false, and ones it gets its own staff or their relatives to write that do the same.

The UK goes a step further. Under the Digital Markets, Competition and Consumers Act, whose rules on this came into force on 6 April 2025, the Competition and Markets Authority's guidance defines a fake review as one that "purports to be, but is not, based on a person's genuine experience", and it warns against picking only the good ones "when these do not reflect the experience being reported by reviewers overall". That applies to a business showing reviews on its own website. A quote that was paid for, given in return for a free product, or written by someone on your staff or a supplier needs to say so.

You can still pick your best quotes. They have to be real, and the set as a whole has to be fair to what customers actually say.

Where to put them#

Next to the claim each one backs up, I think, though I should say I haven't found a study that tests where quotes ought to sit, so this is reasoning rather than evidence. A quote about saving time goes by the line that promises it. One about support goes near the pricing, where people worry about being left on their own once they've paid. A grid of unrelated praise at the foot of the page leaves the visitor to do all that matching themselves, so get your ducks in a row for them.

On the first screen, keep proof to one short line close to what the main button says, either a single named quote or a rating with its source. The longer quotes can wait further down, where someone who's weighing it up will actually read them. A quote from a customer the visitor has heard of is a feather in your cap, by the way, and I'd expect it to do more than a strip of customer logos on its own.

If you'd like an outside view of your own quotes, Magpie's full report checks each one your site shows for the four things above, and reads whether its words are specific and come from a customer rather than from the team behind the product.

Questions people ask#

How many testimonials should I have?#

Enough to back each main claim, and no more than real customers have given you. One specific, named quote does more than a dozen vague ones, so add them as they come in.

Can I edit what a customer said?#

Trimming it for length and fixing a typo is fine, as long as it still says what they meant. What I wouldn't do is add a result they never mentioned, or make it sound keener than they were. And send them the version you're using so they can say yes to it, which is polite anyway and, if it ever comes up, means you can show they agreed to the exact words on your site.

Are video testimonials better?#

They're harder to fake, I think, and there's a real face in them, so they carry trust well. They also ask the visitor for a minute or two. I'd pair each one with a short written line from the same person and a link to the video, so a skimmer gets the point and a careful buyer can still watch the proof.

Sources

  1. Newman, Garry, Bernstein, Kantner and Lindsay, Nonprobative photographs (or words) inflate truthiness (Psychonomic Bulletin and Review, 2012) (opens in a new tab)
  2. Ott, Choi, Cardie and Hancock, Finding deceptive opinion spam by any stretch of the imagination (ACL, 2011) (opens in a new tab)
  3. Nielsen Norman Group, Trustworthiness in Web Design (Aurora Harley, 2016) (opens in a new tab)
  4. Federal Trade Commission, Trade Regulation Rule on the Use of Consumer Reviews and Testimonials, 16 CFR Part 465 (2024) (opens in a new tab)
  5. Competition and Markets Authority, Fake reviews guidance, CMA208 (2025) (opens in a new tab)

About the author

· Founder of Magpie AI

Alex founded Magpie AI and writes about what turns a landing page's visitors into customers, drawing on what its audits find on real sites, from the first screen to the checkout

Related articles

Fix what loses you customers.Fix what loses you customers. Then bring more of them in

  • Free score, no sign‑upNo sign‑up
  • No credit card neededNo card needed