Social Proof: Are You Doing It Wrong?


social proof

[Guest article by Tom van Bommel]

Ever since Robert Cialdini’s Influence became a smash bestseller, his “social proof” principle quickly arose as the world’s favorite influence technique.

That happened for a good reason. After all, what’s an easier way to boost sales and conversions than simply showing what other people have bought or done before?

It’s no surprise that we are all bombarded with social proof appeals. Browsing through the web seems to be an endless stream of product ratings, testimonials, statistics, actions and desires of other people.

I myself was a firm believer of the whole social proof mojo. Needless to say, my heart skipped a beat when I encountered a study last month that changed this for both the worse and the better.

To make a long story short: social proof works, but I was using only a fraction of its full potential all that time.

Social proof helps conversion, but are you doing it wrong? #CRO #psychology Share on X

Liking versus doing

Social proof can come in many forms, ranging from vivid customer cases to dry statistics. However, what Tu and Fishbach (2015) found is that the most game-changing social proof element is not necessarily its execution, but rather how it’s framed.

Social proof can either be framed in terms of actions (“many people bought this product”) or preferences (“many people liked this product”).

Think for a second which type of framing would yield the best compliance results.

If you are like me, your bet is on action-based social proof. I reasoned that the more definitive and concrete kind of social proof would be more influential. And actions are more concrete than preferences, right?

I was wrong.

The researchers discovered that preferences are significantly more influential than actions when it comes to persuasion.

In other words: consumers are more eager to follow what others liked, rather than what they’ve had. This was found across a wide range of contexts, both within the lab and the real world.

When shopping, consumers are swayed more by what other people would like to have, rather than actually have. When dining in a restaurant, people comply with their friends’ choices only when talked about as desires before ordering, but not when the friends’ food is already ordered.

Even diving into the digital depths of YouTube, viewers’ clicks are more influenced by liking scores (i.e., preferences) than views (i.e., behavior). For example, these displays look identical but the second one uses likes rather than views.




To make this principle a bit more vivid, let’s put it in the context of intimate relationships. Would you rather have a partner that’s either desired by many or had by many others? Thought so.

How to boost conversion with preference-based social proof

The superiority of preference-based social proof opens many exciting avenues for conversion testing.

I’m sure you recognize the phrase “other customers bought” or one of its many variations. It would be interesting to see how that converts compared to “other customers liked”.

Similarly, you can rewrite “bestsellers” and “most ordered” in “favorites” and “most wanted.”

Small copywriting alterations like these might just spark a very big difference in conversion.




From the scope of page-structure, the current insights suggest to save a more prominent spot for customer reviews, testimonials and social media metrics (i.e., social proof containing preferences) compared to statistical social proof information such as “bought by more than 10.000 people” (Even that could be subtly rephrased into “enjoyed by more than 10.000 people.”)

As always, the maxim “you should test that” holds true.

The effect wasn’t uniform in all of the experiments, and your situation might change the dynamics. For example, if your company isn’t well known, having 26,450 customers may be more persuasive than being liked by that many people.

There are even rare outlier examples in which social proof reduces the conversion rate. So, run a simple A/B test to see what works for you.

Should you discard action-based social proof?

Absolutely not. If tomorrow’s newspaper exclaims that apples are healthier than bananas, would you stop eating bananas overnight?

Social proof is mankind’s most compelling piece of information. If you can use it, use it. It will help your influence soar, be it rooted in preferences or actions.

But, when you have the option to reframe your message in terms of what people liked instead of what they did, it’s likely your message will pack an extra persuasive punch.

Study: preference-based social proof can work better than action-based. #CRO Share on X

  1. Amy says

    While preference is far more persuasive than purchasing, in our (online) world one can “like” something without having intimate knowledge of or purchasing. Example – I can like one of the cat videos referenced in the article based on the still of the video alone, without actually watching any of it. When marketing, clarity of message is so important, and likely is one example where that messaging can quickly get misconstrued.

    1. Tom van Bommel says

      Thanks Amy! Certainly. As liking is often such a small step (and sometimes meaningless, such as liking a cat video you didn’t even watch), I initially expected it to be outperformed by ‘doing’. But interestingly, it seems people don’t take this into account – which makes sense considering the often irrational and automatic way we process information.

  2. Chester Butler says

    Nice article, Tom. The liking concept has been around a long time. There is an old proverb that says,” Before you start a business, learn to smile.” Here’s my take, social proof, like or action oriented, is easily negated by dealing with a person who is not likable. Social proof is just the beginning of a relationship.It is not a relationship.I’ve never met a successful sales person that could not turn on the charm while building trust.Social proof is the fuse, liking is the bomb that changes behavior. Boom, Boom, Boom!

    1. Tom van Bommel says

      Thanks for your comment Chester! I expect it to hold the other way as well: a likable salesperson will certainly generate more social proof (e.g., word of mouth) than someone obnoxious 😉

  3. Edmond says

    Thanks Tom for the informative post. I had never heard about social proof before. Reading the article I noticed how social proof influence my buying decisions. In have an information blog, I write stuff for people to read. I was wondering how can I use social proofing to increase readership of my blog. I will be glad if you can give me some tips.

    1. Tom van Bommel says

      Thanks for the compliments, Edmond. Social proof in the blogosphere boils down to showing your readers that others like your posts. The easiest way to accomplish this, is simply showing the well known social liking and sharing statistics. But with some creativity, there are certainly more methods to juice up the social proof. Good luck with your blog!

  4. Bren Murphy says

    Hi Tom,
    I am always on the look out for new ways to optimize and tweak my presentations and this has truly provoked me to assess where my bias lies. For sure actions are good, but to suggest that preferences are more powerful changes the way things will be done for me. I’m not entirely convinced though – and might split test some of the variations to see how they play out for me in my setting.

    1. Tom van Bommel says

      Hi Bren, split testing both types of social proof would be very interesting! Please keep me posted on your findings.

  5. Joost Fromberg says

    Great article Tom, let’s hope will get even more subscribers after this excellent blog!

  6. Steve Beazer says

    Thanks for the article. I appreciated the statement “irrational and automatic way we process information.” in the comments above. It is an interesting world we live in.

  7. Umair says

    Liking maybe a metric but personally i would love to dwell deeper into the numbers presented.

    Someone may get influenced by the likes but is the resulting ‘Click’ e.g. tangible (e.g. did he actually see the complete video) or did the people who clicked or saw the likes, actually purchase more and by how much.

    We ourselves click based on how much people are sharing / liking but the real metrics are still revenue driven. Perhaps, a deeper study might shed more light on this.

    1. Tom van Bommel says

      I’m also eager to see more research measuring actual spend dollars. You might want to look further in the full article though (, as many of the studies do involve some form of real world choices (e.g., choosing a meal, shopping on Amazon).

  8. Neil Spencer says

    We recognize that a certain item is desired by many and therefore it becomes valuable simply because it is desired.

    It would be interesting to discover why then some people are motivated by this social proof to just desire the object like they see others doing (by hitting “like”) and why some go that extra step and purchase it.

    1. Craig Atkins says

      Groupthink, “Keeping up with the Joneses”, etc… there are many explanations, but really they all boil down to the fact that humans are social animals and thrive off of being a part of a group.

  9. Jono says

    Tom, do you have any thoughts on how the concept of “liked” versus “did” might apply to the world of nonprofit marketing and fundraising?

    1. Roger Dooley says

      It seems like the same basic principle would apply, Jono… For example, it would be better to show how many people donated to a cause than signed a petition in favor of it. Every situation will be different, of course.

  10. Nia says

    But how about people’s response over some survey whether they liked the food/product and say yes, but would actually not repeat the purchase. Wouldn’t action speak louder then? That’s how social media generate many likes (thumbs up) but has been criticised to not generate enough action. In that case, how do you think I can convert a “reason to believe” (preference) to “reason to buy” (action)?

    1. Tom van Bommel says

      Hi Nia, certainly! When people talk about themselves, their action intentions probably predict behavior better than mere liking. This is the classic consistency principle coined by Cialdini: people that verbally commit to an action are more likely to stick to their intentions. A great example of this phenomenon is that marketing surveys that ask people about their repeat buying intentions actually increses repeat purchase!

  11. Nick Marshall says

    Perhaps the Wish List that many shopping sites have could be put to better use too. Something like a monthly top 10 wish list perhaps.

Leave A Reply

Your email address will not be published.