Blog
Business

Social proof helps only when it says something

During a few days in Paris in 2012, I saw street musicians with a few coins already in their hats. I thought that small sign might show passersby that others had given something too. The idea of social proof appears on websites as well.

Online, though, it can easily become a collection of tricks. A counter says people are viewing a page, a badge gives a customer count, and stars appear beside a product. The question is not how many signals you can display. It is whether a visitor learns anything useful from them.

Show what is relevant

A review is most useful when it is clear what it describes. Someone looking for an online shop may get more from feedback by a person with a similar need than from a general compliment. A client case study can provide more context: what was the assignment, which decisions were made, and what came out of it?

Feedback can also help with a product. A review explaining how someone uses it often says more than a star count alone. Leave room for nuance. If every response is glowing and interchangeable, it becomes hard to assess them.

In 2012, my then online shop Groovy Gadgets showed items other customers had also bought in the basket. That offered a more concrete suggestion than a stand-alone counter, although I could not tell from that alone whether customers found it useful or bought more because of it.

Numbers without context are weak

Back then, I listed many possibilities: bestsellers, viewer counts, recent purchases, and stock levels. Some can be useful information, but only if the data is accurate and relevant to the visitor's decision. “Three left in stock” helps someone who wants to know whether they can order. A random counter designed to create urgency is less helpful to me.

So check which question a number answers. Is the time period clear? Is the source clear? Is the information current? If you cannot explain it, leave the number out. A number on a page does not automatically create trust.

An empty counter says something too

In 2014, I wrote about share buttons beside blog posts. On my own blog at the time, the numbers were often modest, and I removed the share buttons. A zero next to an article may distract readers, but I cannot prove it actually stopped anyone from reading. The main point is that such buttons take up space without always adding value.

If you want to make sharing easy, check whether a button is actually used in your situation. The share count does not have to be visible. I hope readers judge an article by its content; a low count is not a measure of quality. A high count does not improve a weak article either.

Use proof to explain

For every signal, ask: what can a visitor judge from this? A specific review, a fitting example, or an honest list of frequently bought products may help. An unchecked badge or counter may raise questions instead. If you show evidence, explain what it represents and keep it accurate.

The coins in that Parisian hat made for an interesting observation. On a website, I find an honest account of a real experience more useful than a whole row of clever nudges.

— GijsDiscuss this article