In an experiment conducted more than thirty years ago, volunteers were presented with two bowls of different sizes. The large bowl contained ten red jelly beans mixed with 90 white jelly beans. The small bowl held just one red bean and nine white beans. If the volunteers closed their eyes and picked the red bean from one of the bowls, they’d win four dollars. The bean selection would be random but the volunteers could choose the bowl.
Almost 77 percent chose to hunt for the red jelly bean in the large bowl.
The decision was entirely irrational. The odds of picking correctly were one in ten regardless of which bowl the volunteers used. But the volunteers had focused not on the probability of winning their lucky dip but on the number of red beans available.
They’d exhibited ratio bias, a propensity to perceive a low-probability event as more likely when its probability is represented by a large number in the ratio than by a small number, even when the probability is equal or even higher. In practice, sometimes people prefer to try their luck at odds of two in thirty than one in ten because two looks bigger than one.
That apparent irrationality has important implications for marketers. Sometimes, what appears to be the most rational way of presenting information isn’t the most persuasive.
That’s especially true when presenting popularity cues.
If All These People Like It, You Will Too
Popularity cues are indicators of the popularity of a product or service. When potential buyers are unsure about the quality of a product, they’ll often follow the actions of others. If other customers have made a product a top-seller then that product is likely be of high quality. Shopping sites like Amazon list category “top sellers” and “popular products” to reduce what could otherwise be a bewildering range of choices to a manageable selection of what appears to be the best options.
Popularity cues can also indicate scarcity, a technique often employed in the tourism industry. Travelers looking for hotel rooms at Booking.com, for example, are told how many of each type of room remain available. Because the numbers are frequently low—or appear low—customers feel pressure to decide quickly before their options become limited.
But some outlets provide even more precise social proof. As users browse the site, they’re told how many people purchased the product or service over a set period of time. Disney, for example, has been known to promote its resorts by telling potential customers how many bookings the resort has taken over the last 24 hours. By combining the total number of sales with a time measurement, customers can see the rate at which a product is selling as well as its popularity. They get both the social proof that comes with the impression of a popular product and the sense of scarcity as the speed of sales suggests decreasing availability.
Those sales messages combine two elements: the volume of sales represented by the number of bookings; and temporal information represented by the period in which those sales were taken. But those two pieces of information present marketers with a difficult dilemma. In theory, the faster a product is seen to sell, the more pressure customers will feel to make a purchase.
But we’ve seen that people don’t always calculate ratios carefully. Should marketers focus on presenting large numbers of total sales over a long sales period? Or should they emphasize the sales rate by reducing the sales period at the expense of the number of sales?
“It is possible that people might ignore temporal information and compare total sales,” say Bi Yang, Shanshi Li, and Zhenyu Chen of the School of Management at China’s Xiamen University, and Anna S. Mattila of the School of Hospitality Management at The Pennsylvania State University in a recent study. “On the other hand, it is also possible that when evaluating hotel popularity, individuals are inclined to take temporal information into account and compare hourly sales.”
So what’s the most effective way to present social proof of online sales?
The researchers conducted a series of experiments to help marketers refine their presentations of real-time sales data by comparing the effect of displaying large total sales over a long period to large hourly sales over a short period.
The researchers also looked at who is making those evaluations. They focused on experiential thinking—cognitive shortcuts that simplify complex judgments in order to make fast decisions, such as the ratio bias has people picking beans from a larger bowl. The researchers assumed that people who often engage in experiential thinking would be more likely to focus on the hourly rate than on the total number of sales.
First, the researchers asked 85 volunteers to imagine they were about to take a vacation with a friend, and were considering booking a hotel room. The volunteers were told that one hotel had received 72 bookings in the last 24 hours; the second hotel had received 24 bookings in the last six hours. When the researchers asked which hotel was most popular, just over 22 percent chose the first hotel. A repeat of the experiment with a larger gap between the booking rates lowered the choice to a little over 18 percent but the results still showed that around in five people were susceptible to ratio bias. They were looking at the total number of sales, not the rate of sales.
A second study added questions used to measure experiential thinking. Volunteers were asked to indicate whether they agreed with statements such as “I trust my initial feelings about people.” The results suggested a correlation between experiential thinking and ratio bias. A one-unit increase in a participant’s experiential thinking score was associated with a 78 percent increase in the probability of choosing a hotel that had been booked more times than a hotel that was being booked at a faster rate.
The researcher had showed that the more people look for shortcuts during their decision-making, the more likely they are to be influenced by the first big number they see. That bias was reinforced in the findings of the last experiment. In this trial, the researchers reversed the presentation of the booking rate. Instead of telling the volunteers that the hotel was “booked 72 times in the last 24 hours,” they wrote “In the last 24 hours: Booked 72 times.”
People who scored low in experiential thinking—people who were less likely to look for shortcuts in their decision-making—were now even less likely to choose the hotel with the higher total bookings than the hotel with the higher booking rate.
“In contrast, for people high in experiential thinking style, presentation order did not significantly influence their hotel booking decision,” the researchers concluded.
So what does this mean for marketers? How should websites use sales data and the frequency of sales to nudge visitors towards conversions?
Display Sales by the Hour
The researchers noted that the study compared two types of sales messages: large total sales over the long term compared with large hourly sales over the short term. “Our findings reveal a weak ratio bias effect in the hotel booking context,” they concluded. “That is, a small number of people believe that large total (vs. hourly) sales indicate greater popularity. Such people tend to rely on the experiential thinking style to make judgments.”
They also found that displaying the time period before the sales volume reduced that ratio bias but only for people who don’t take short cuts in their thinking.
Marketers, they noted, tend to communicate and display product sales over the long term in order to cash in on an intuitive belief that greater sales imply higher product popularity. However, because only around one potential customer in five uses ratio basis to make decisions, marketers can expect to see better results by focusing on the rate of sales.
“Our findings indicate that most people tend to evaluate product popularity based on hourly sales,” they conclude. “Thus, online booking platforms should apply algorithms to calculate and compare hourly sales within certain periods of time (e.g., 3, 6, 12, and 24 h) and display the sales message with the largest hourly sales volume.”
But it is also possible to target the shortcut-thinkers who are more likely to be moved by larger figures. The researchers argue that young people and women show higher rates of experiential thinking, and that this thinking style is also more likely to be applied when using touchscreen devices such as mobile phones and tablets.
So marketers targeting young or female audiences, or whose sales pages appear on mobile phones, should experiment with presenting sales data that use large number totals over long periods rather than high rates of sales over short periods.
Recent Comments