× Patterns Blog BetterUI Review Coaching Courses Datastories

Evidence #71: Social Proof That Didn’t Work

Tagged As: Result

Thanks to Daniel Bridges from findsomeone.co.nz for sharing Evidence Test #71. This New Zealand’s dating site test suggests that sometimes social proof (in this case profile photos) might actually backfire. (Note: we ran this test over 2 years ago).

What Can We Learn From This Test?

The variation with the profile photos on the homepage showed a possible -6.6% decrease to searches. We aimed to increase searches as an interim metric to signups (if people search more, they should find more interesting profiles, and have a reason to signup – which is required for making contact). Here are some thoughts as to why this test went slightly south:

Share this story

Posted by Jakub Linowski on Feb 2, 2017

Leave Your Reply

10 Comment threads
2 Thread replies
Most reacted comment
Hottest comment thread
11 Comment authors
TimJakub LinowskiLisanneAshSheldon Gay Recent comment authors
newest oldest most voted

I also think showing the photos could increase worries on privacy and provoke questions like “Will my photo show up here when I sign up”. Questions which can lead further down the path of worrying about sharing personal info online.

Darek Markiewicz

I am assuming the generic profiles were selected randomly. This will cause unanticipated focal points to show up and grab attention, for example the highest contrast image (dude in a hat on a white background in the middleleft) is the top image in the visual hierarchy of just the profile photos. Well I happened to be a heterosexual guy, so a photo of another male does not appeal to me on a dating site, it’s not my target it’s my competition. Just a thought, which is really variation on “Diluted Attention” but more specifically pointing my attention in a wrong… Read more »

Charlie Hathaway

The banner image at the top of the page is quite emotive, drawing users in the ‘idyllic relationship’ whereas the profile images are a stark opposite of this. They lack the element of fun and excitement that is seen in the above image as often the profile images are just plain, ordinary head shots. In a way, the site seems ‘less professional’ with the profile pictures owing to the lack of consistency between the images – seeing them stacked in search results, this would be much less evident and more expected. Furthermore, you’d also hope that the images used are… Read more »


Interesting. My first thought was that the difference is so low that it’s almost negligible. I’d like to see this test carried out over a longer period of time to account for any anomalies. But if I were to indulge more in this, I would guess along the following lines: – social proof tends to have more influence during instances where people are uncertain about making a decision, and see similarities in others who are taking or have taken an action in a similar instance. – if we put ourselves in the shoes of a prospective user who is deciding… Read more »

Loren O\'Laughlin

The provided photos may be a poor match for those who are turned away. I know it’s supposed to be “social proof” but in this context it’s actually a preview of the product… kinda like restaurants that put bad photos of food on big menu boards—instant turn off.

Ty Cahill

Doesn’t look like social proof to me. Just looks like a bunch of low quality head-shots that don’t even match the style of the website. Are these real people? Nothing tells me who they are or where they’re located. All I know if they’re “recently online.”

Sheldon Gay
Sheldon Gay

I think I would add to your thoughts on “distraction” and “too general” the fact that these faces aren’t really “social-proof”. They aren’t examples of happy daters. They’re a bunch of random faces that the user may or may not be interested in and is probably at least a little confused with since it has multi-gendered examples. Most users we could assume are primarily heterosexual and therefore aren’t looking for pictures of both genders at any given time. If these were true “social-proof” examples instead, the multi-gendered photos would be acceptable because it would be conveying the idea that many… Read more »


Just to add to this. Social proof is working -> just not in the way intended.

The one thing that we can take away is that social proof negatively impacts conversion. I would suggest following up with a test removing other social proof or profile photo instances on the website to check the Elasticity of social proof. It seems the elasticity of that change is high so why not try removing instances of example pictures or other forms of social proof. It may increase conversion since we now know that visitors are turned off by it!


My guess: it may trigger the question “what happens with my image when i sign up?” or visitors may think “Does my profile image show up on this homepage when i sign up myself?”. I can imagine that in dating market, privacy is a thing. For sure I would run a usability test on this concept before development/launch.


Here’s an interesting thought. Based on this article (see below for link), scarcity (I know, not social proof, but I will get there), has two ways of showing itself. Either by ‘almost sold out’ and ‘limited edition’. It turns out, only for people who have a ‘high need for uniqueness’ the ‘limited edition’ works. They figure it’s because if you want to be unique, you want something no one else has. This of course comes back to some kind of negative social proof. In a way that ‘almost sold out’ is also a sign of social proof (a lot of… Read more »