Survivorship bias in testimonials: why your best customer results can scare off savvy buyers

On a coaching, training or info-product landing page, the temptation is strong to open the testimonials block with the most spectacular result available: "$0 to $10,000 in revenue in six weeks," "-20 lbs in two months," "I doubled my leads in ten days." That testimonial exists, it's true, and it's tempting to lead with it. The problem isn't that it's false — it's that it's unrepresentative, and a sufficiently savvy visitor — often the most qualified prospect, the one who's already been burned by a similar offer — perceives it as such. It's a textbook case of what statisticians call survivorship bias.

What survivorship bias actually is

The name traces back to a specific episode from World War II, documented by statisticians Marc Mangel and Francisco Samaniego in a 1984 article published in the Journal of the American Statistical Association (see on Google Scholar). The US military wanted to reinforce bomber armor by studying bullet-hole patterns on planes that returned from missions, in order to armor the most-hit areas. Statistician Abraham Wald flipped the reasoning: the impacts observed on surviving aircraft showed exactly the areas a plane could take damage in and still make it back. The truly critical zones were the ones with no visible damage — because planes hit there never returned. The observed sample (the survivors) gave an inverted picture of reality.

On a landing page, the mechanism is identical: the testimonials shown are the "survivors" of a selection process that's usually invisible to the visitor — the customers who got the best result, under the best conditions, with the most personal effort invested. Customers who followed the same method with a modest result, or none at all, are almost never shown. The visitor only sees the top tail of the distribution and wrongly infers that it represents the average.

Why this ends up costing conversions, not just an ethics problem

One might assume a spectacular result, even an atypical one, can only fuel the desire to buy. That's true for part of the audience — the part that hasn't yet been burned by similar promises. But research on advertising skepticism shows this audience isn't homogeneous. Marketing psychologists Carl Obermiller and Eric Spangenberg developed and validated a scale measuring consumer skepticism toward advertising in 1998, in the Journal of Consumer Psychology (see on Google Scholar): this skepticism is a stable trait that varies across individuals, and it activates precisely in response to claims perceived as exaggerated or unverifiable. The buyers most qualified for a paid coaching program or course — the ones who've already invested in competing offers, who compare several pages before deciding — statistically score highest on advertising skepticism, not lowest. The extreme testimonial, meant to convince them, is exactly the kind of signal that triggers their guard.

What research shows about vague disclaimers

The classic instinct for covering yourself is to add small print: "results may vary" or "results not typical." A study commissioned by the US Federal Trade Commission, led by marketing researchers Manoj Hastak and Michael Mazis and published in 2004, tested the effect of this kind of disclosure on consumer perception (report available on ftc.gov). The core finding: a vague disclaimer like "results may vary" barely corrects the impression left by an extreme testimonial — participants kept overestimating the result they could expect for themselves. A concrete, specific disclosure ("the typical customer achieves this result, over this timeframe"), on the other hand, significantly reduced that overestimation. The difference isn't whether a disclaimer is present, but how precise it is: a number against a number, not a number against a lawyer's phrase.

What this means for LanderKit's French-speaking audience

In France, the advertising self-regulatory body ARPP requires testimonials in advertising to be authentic, verifiable and relevant to the claim being made, and bars them from becoming misleading through an exceptional or outdated character. For a coach, a training provider or an info-product seller, this maps directly onto the survivorship-bias problem: the testimonial itself isn't the issue — presenting it as a representative outcome is, both from a compliance standpoint and from a real conversion standpoint.

How to show strong results without survivorship bias

• Pair the exceptional result with a typical one: "Léa doubled her sign-ups in three weeks; most clients see a 15-25% lift in the first month" informs as much as it impresses.
• Contextualize the number instead of isolating it: mentioning the starting point, the time invested and the real timeframe turns a result that looked like magic into one that looks achievable — and therefore credible.
• Show several result profiles, not just the best one: a customer who improved slowly but genuinely often reassures more than an outlier, especially for a buyer who doubts their own ability to succeed.
• State a verifiable range or average instead of a vague "results may vary" — that's the disclosure that, per the research above, actually corrects perception without weakening the message.
• Date testimonials and refresh them: an exceptional result achieved three years ago, under an offer that has since changed, is just as misleading as an unrepresentative result was at the time it happened.

What this doesn't mean

This principle isn't a license to drop your best testimonials or dilute them into a bland average — a striking result remains a powerful argument, as long as it isn't presented as the norm. It connects to what we cover in our article on the imperfection effect: a social-proof signal that's too smooth, too perfect, or too extreme raises suspicion rather than trust. For the structure of a longer testimonial that sets this context without burying it, our guide on the customer case study details how to tell a result in a verifiable way rather than displaying it as an isolated number. And for how many testimonials to show in the first place, see our article on how many testimonials to display on a landing page.

The 10 LanderKit templates (€89 each, €229 for the full bundle) leave complete freedom over the testimonials block's content, with no forced number or staging — whether for a coaching offer, a certified training program or an ebook or info-product. You can see the block in context on the coach-consultant demo before adapting it to your own results, best and typical alike.