When 8/10 Doesn’t Equal 80/100

Numbers are often used in ratings and evaluation of products and services. But how numbers are presented can have a big influence in the eyes and minds of consumers, and can make or break a sale says recent research.

Does 8/10 equal 80/100? The answer seems obvious as both of them equal 80 per cent. But you may want to think again after reading a recent research jointly conducted by the Chinese University of Hong Kong (CUHK) Business School and Tsinghua University School of Economics and Management.

The research was conducted Tao Tao, a PhD student at the Department of Marketing of CUHK Business School, in collaboration with Prof. Robert S Wyer and Prof. Yuhuang Zheng, faculty members at CUHK Business School and Tsinghua University, respectively.

The team found that consumers would perceive a product rated as 8 out of 10 better than a product rated 80 out of 100, whereas a product rated as 2 out of 10 is perceived as worse than a product rated 20 out of 100.

The Popularity of Review Scores

Nowadays, consumers often see numbers and scores in marketing brochures or the Internet. Before making a purchase of a certain product or service, they may search on the web for the reviews. If they want to see a movie, they will visit a movie review website to find out about the score. Likewise, they will go on a restaurant review website to see how past customers rated the food and experience of dining there. Very often, consumers will make choices based on the scores of satisfaction given by other customers.

Consumers perceive a product rated as 8 out of 10 better than a product rated 80 out of 100

Marketers may also use the same tactic to demonstrate the performance or popularity of a product through numbers and scores, which are straight forward and easy to understand. For example, marketers might advertise that 7 out of 10 people use their products, or that their product’s quality performance obtains a score of 88 out of 100.

However, the range of the scale along which scores are defined could be either small or large. For example, the satisfaction rating of a film could be expressed as 7.8 out of 10 or as 78 out of 100. The market share of a product could be presented as 7 out of 10 or as 70 out of 100. Mathematically, they mean the same thing; however, they differ in their influence on consumers’ judgments.

The Scale Range Effect

In the research, the authors found that when people encounter a score expressed along a bounded score, that is, 9 out of 10 or 90 out of 100, they would initially determine whether the score is good or bad.

If they can clearly categorize the score as a good or bad one, this initial impression would form their final judgment of the score. In this case, the evaluation of the product would not differ between 9 out of 10 or 90 out of 100.

However, if they are unable to confidently categorize the score as good or bad, this ambiguity would push them to further consider how good or how bad the score is.

Here is where it gets interesting. If the score is initially categorized as a moderately good one, people would further compare the score with the high endpoint, that is, the maximum score. For example, 8 on 10 would be compared to 10 and 80 on 100 would be compared with 100)

However, if the score is initially categorized as a moderately bad one, people would compare it with the low endpoint, that is, the minimum score. So 2 on 10 score and a 20/100 score would both be compared to a 0 score.

Because people have the tendency to perceive a large number to have more value than a small number without considering its unit, the discrepancy of 20 would be judged as larger than a discrepancy of 2.

Therefore, 8 out of 10 would be evaluated as better than 80 out of 100 since the gap between 8 and the maximum score, 10, is perceived to be smaller than that between 80 and 100. In contrast, 2 out of 10 would be evaluated as worse than 20 out of 100 since 20 is much further away from the minimum score level, 0, than 2.

Using mainland Chinese, Hong Kong students and U.S. citizens as samples, the authors examined the effect scale range in different scenarios, including restaurant and product satisfaction ratings, teaching evaluation scores, examination scores, and the voting process for politicians.

When a book sold on Amazon was rated as 7 out of 10, consumers would give a higher evaluation and were more likely to purchase it than when the book was rated as 70 out of 100

They found that when a book sold on Amazon was rated as 7 out of 10, consumers would give a higher evaluation and were more likely to purchase it than when the book was rated as 70 out of 100. For a poorly reviewed product rated 3 out of 10, consumers would perceive it as worse than one with a rating 30 out of 100 and therefore be less likely to purchase it.

When it comes to voting for the next president, if people were considering the supporters of the candidate, they would perceive the voting rate as higher when presented as 82,080 thousand out of 171,000 thousand rather than as 82.08 million out of 171 million. However, if they were considering the citizens who did not support the president candidate, they would perceive the former voting rate as lower than the latter, and thus be less likely to support the candidate.

Numerical information of the sort described above can strongly influence consumers’ decision making. The research may provide a useful guide for marketers who want to convey product information in a way that will maximize its impact.

For example, they might choose a relatively large range scale if the target rating score is low, thus increasing its numerical discrepancy from the low scale endpoint. However, they might want to choose a small range scale if the target rating score is high, decreasing the score’s discrepancy from the high scale endpoint.

Authored by Tao Tao. Tao is a PhD student in the Department of Marketing at CUHK Business School. Article courtesy CUHK.