The 25 Percent Rule

Social norms can be changed by only a quarter of the group

Whether it's 10 or 10,000 people, changing the behavior, beliefs or norms of an entire group is hard, but new research suggests that in order to do so, we actually only need to convince 25% to start a persuasive domino effect.

Centola, D., Becker, J., Brackbill, D., & Baronchelli, A. (2018). Experimental evidence for tipping points in social convention. Science, 360(6393), 1116-1119.

Imagine this: you manage the workplace of a mid-sized creative agency in London. The team are always rushed off their feet with deadlines and, being in London, lots of tea gets drunk every day.

As a result, in the midst of each frantic day, people never clear their cups away; the place is a never-ending mess. When clients come, you're always embarrassed at the state of the place as you just know it has an impact on their perceptions of professionalism.

So what do you do? How can you get the team to shift their behavior and tidy a little more?

Initiating behavior change amongst a group is no easy feat. However, the effort required to do so needn't be as great as you might think.

New research suggests that in order for a group to adopt a new behavior, only 25% of the group need to change in order to influence and convert the rest.

This is the first time we've had quantifiable evidence of a 'social tipping point'. So, if it holds, it's incredibly exciting, and will have a profound impact on how we affect social change with regards to all sorts of norms, be they in the workplace or in a cultural, political or consumer context.

Let's take a look at the research.

The study

Making use of an existing model for exploring the emergence of social conventions, the study was split into two phases: the first sought to create a group consensus and the second to override it.

Phase One - Creating group consensus

194 people were randomly assigned to one of ten online groups and interacted in a name game.

In each round of the game, two randomly-paired players simultaneously assigned names to a pictured face. They'd be awarded $0.10 each time they wrote the same name and penalised the same if they differed. 

After each round, players could see the choices they and their partner had made to discover whether they matched. A record of their match success rate was viewable at all times. 

This interaction process quickly led to the establishment of a group-wide consensus as players began suggesting the name that had won them the most amount of matches.

Phase Two - Overriding group consensus

The aim now was to see whether existing group-wide opinion on the name could be changed with a committed minority group seeking to rename the face.

Ten trials were conducted: five composing of a range of minority groups below 25% and five composing of a range of minority groups over 25%.


The results showed that efforts by the minorities less than 25% weren't effective at changing the names, with only 6% of the larger group adopting the alternative name.

However, when a larger committed minority was used (between 25-31%), the new alternative name became the norm.

Interestingly, in one of the groups, just increasing the size of the minority by one person lead what was otherwise a failure to a successful group behavior change.

Important to note is that the dynamics of the game sought to reflect how linguistic and behavioral conventions emerge during online communication.


⚠️ In the research, social and financial incentives were awarded for co-operation, meaning that even selfish people saw the benefits of change. How might they act instead if the benefits of doing so were either not so clear (unquantifiable gains) or the benefits were only gained at some future point (e.g. halting sea level rise), rendering them less attractive due to our Present Bias and tendency to hyperbolically-discount such things. 

⚠️ More research is needed to explore how the confirmation bias, sunk cost bias or the emotional need for consistency with our prior decisions affects an individual's resistance to group change.

⚠️ Though it is robust in the findings and held up in simulations between 100 and 100,000 people, it's safer to assume 25% to be a benchmark rather than a universal value. Clearly, it will differ according to specific context (historically-entrenched beliefs etc) and should be tested accordingly.

Concluding thoughts

For better or worse, any behavior or belief can be considered 'normal' if held by enough people in a group. So whether you're just looking for a cleaner workplace or are trying to change a group's acceptability of an activity like smoking or bullying, it can seem like getting others to change is a seemingly impossible act.

But what this research shows is that we have more power to change these norms than we think. Assembling a small, determined quarter of the group appears to be your first step.

Key takeaways

Determine the specific social change you wish to make, ensuring that it's unambiguous, measurable and has a visible feedback loop for the group.

✅ Highlight the wider, strategic importance of the change with an understandable narrative that ties with the company's vision:

✅ Enlist a subgroup of 'Change Champions', roughly 25% of the group to instigate the new norm. You may consider attaching some subtle status or career-orientated reward to these individuals.

Determine program length. Consider running the behavior change program for a specific period of time in order to account for sufficient opportunities for change. The trials in the research were run for around 4 months, for instance.

Ensure commitment. For the length of the program, ensure a consistent level of commitment from these individuals. Commitment to the change tends to matter much more than the authority carrying it out (Xie et al, 2011). For instance, it's vital that Change Champions take their cups out of the meeting room each and every time.

Create group-wide incentives. You may also consider some group-wide incentive mechanism to reinforce the change, especially once adoption of the new norm reaches a certain measurable threshold.

Communicate the impact. Close the loop on the new social norm by celebrating the positive impact of the change with the existing group. Communicate it clearly to any new members, along with the underlying strategic reason for its importance.  

What’s a Rich Text element?

The rich text element allows you to create and format headings, paragraphs, blockquotes, images, and video all in one place instead of having to add and format them individually. Just double-click and easily create content.

Static and dynamic content editing

  • Aspirational membership schemes and belonging The category size bias provides a credible explanation for why we human beings tend to associate with large groups that are viewed favourably by society. Being part of a large and “desirable” social group can make others believe that we also possess the many qualities of its members. For small businesses, it suggests that forming or being a part of a consortium or large and high quality networking group can dramatically elevate your brand image.
  • Communicating category sizes to nudge effectively Highlighting the differences between the large and small categories is highly likely to enhance the effect of the Category Size Bias. For instance, for software companies, stating that there are 10 features in the premium version versus 4 in the free version will help nudge a decision towards the premium version

How to customize formatting for each rich text

  1. The findings from this braingem can nudge better healthcare choices, encourage consumption of a given product, and lead to more confident consumer decisions.
  2. We mistakenly believe that items in larger categories have a higher probability of being picked than ones in smaller categories, despite all items having an equal chance of being picked.
  3. We’ll spend or gamble more money on items put in larger categories.
  4. We’re more likely to take action from tasks when they’re in a bigger list, over a smaller list.
  5. We once we put something into a group, we perceive it to adopt all the characteristics of that group. This suggests that small companies should foster alliances with similarly-principled, more established companies.
Back to all Research