Which social proof examples actually lift conversion?

Social proof lifts conversion when it answers a doubt the visitor has at that moment. In our A/B tests, a monthly count of people who signed up, a row of media logos and a customer service award all won. A count of how many people viewed a product lost on every device. The format was not the difference. What mattered was whether the proof answered a real question.

Below are 6 social proof examples we tested, what won, what lost, and where each one belongs on your site.

What social proof is

Social proof is evidence that other people chose, used or approved of what you sell. It shows up as reviews and star ratings, counts of customers, logos of media or partners, awards, and testimonials. It works because people look to others when they are unsure. The key word is unsure. Proof that arrives before there is a doubt, or after the doubt is gone, does little.

A monthly count lifted lead form conversion 21.6%

Our first social proof test on 1 lead form lost. We came back with a larger number and tested it 2 ways: counted for today, and counted for this month. The monthly count won. Over 29 days, conversion rose 21.6% and value per visitor rose 20.8%.

Evidence · Lead form · Won

Does a monthly social proof count lift lead form conversion?

29 days · responsive · We tested a larger social proof count, shown daily or monthly.

+21.6% conversion, +20.8% lead value per visitor

Read the full test →

The bigger number did the work. A daily count can look small early in the day, while a monthly count shows real scale. When you pick a time frame for a count, pick the one that shows how many people actually act.

2 more details from this test are worth copying. First, we came back after a loss. The first social proof test on this form lost, and the second, with a larger number, won. A loss tells you the element matters, not that the idea is dead. Second, the result moved in the right direction for several weeks before it had enough data to call. We let it run the full 29 days rather than calling it early.

Media logos lifted buying guide conversion 47%

A buying guide that sent readers to a retailer showed a review team section but no outside proof. We added an “as seen on” row of well-known media and social platform logos. Over 22 days, click-through to the retailer rose 6.3% and conversion rose 47%.

Evidence · Article page · Won

Do as seen on logos lift buying guide conversion?

22 days · responsive · We added an as seen on row of media logos to the buying guide.

+47.0% conversion, +6.3% click-through rate

Read the full test →

The review team section told readers who tested the products. The logo row told them other people trusted that team. Outside recognition made readers more ready to buy.

3 more wins from outside proof beside the product

The same pattern showed up on product pages:

• Brand logos beside a lifestyle image lifted desktop conversion 20.4% and revenue per visitor 22% (see the test).
• A customer service award badge in the product gallery lifted mobile conversion 12.1% and revenue per visitor 31.8% (see the test). The same badge lost on desktop, where the page already had enough trust signals.
• A monthly units sold count, paired with shorter value props, lifted mobile conversion 13.7% (see the test).

We cover where these fit on a product page in Shopify product page optimization: 6 steps from real A/B tests

Shoppers preferred the page without a view count, by up to 35.0%

Not all social proof helps. We added a count of how many people viewed each product, tested by week and by month. Shoppers preferred the page without it. Conversion fell on every device and version, by 11.5% to 35.0%. Mobile took the bigger hit: conversion fell 35.0% with the monthly count, and revenue per visitor fell 34.2% with the weekly count.

Evidence · Product page · Lost

Does a social proof view count on the product page lift conversion?

14 days · responsive · We added a count of how many people viewed the product, by week or month.

-35.0% conversion, -34.2% mobile rpv, var a

Read the full test →

A view count proves people looked, not that they bought or liked it. It can even raise a new doubt: if so many people looked, why did they not buy? Proof should show people acting, not browsing.

One detail I would keep even though this test lost: we showed exact view counts, not rounded ones. A number like 19,536 reads as real, while 20,000 or 20k looks like someone estimated it or made it up. If you test social proof numbers, use the real ones.

How to test social proof on your own site

Social proof is 1 of the easiest things to test, because it rarely needs new pages. A few rules we follow:

• Write it as a hypothesis. By adding a monthly count of sign-ups above the form, we will reassure hesitant visitors, resulting in a higher completion rate.
• Test 1 type of proof at a time. If you add logos, reviews and a count together and the test wins, you will not know which one to keep.
• Read mobile and desktop separately. The award badge won on mobile and lost on desktop. A blended result would have hidden both.
• Use real, exact numbers. Never round or pad a count to make it look bigger.
• Run at least 2 full weeks. Proof can take time to show its effect, as the monthly count did.

Where to put social proof on your website

Put each piece of proof next to the doubt it answers:

• Homepage and landing pages: outside recognition, such as media logos or awards, to answer “can I trust this brand?”
• Product pages: star ratings near the title, and outside proof near the image or price, to answer “is this the right product?”
• Lead forms: a count of people who signed up, in the time frame that shows the most people acting, to answer “do people like me do this?”
• Checkout: guarantees and service proof, to answer “what if something goes wrong?”

Then test it. As the award badge showed, the same proof can win on mobile and lose on desktop. For how social proof fits into a wider plan, see What is ecommerce conversion rate optimization?

Social proof works when it shows people acting, not people looking.