The Left-Digit Effect: Not All Pennies are Created Equal

The impact on expensiveness perceptions by pricing on the 9s versus the 10s

"Should I offer my product at $3.49 or $3.50?" This was a question posed by a client who was struggling to control the expensiveness perception of her brand. Research had consistently shown that consumers had rated this client's brand as the most expensive in the marketplace and had identified this expensiveness as a significant purchase barrier. The client needed to increase prices and wanted to understand whether to offer the price at $3.49 or $3.50.

Research into the Left-Digit Effect suggests that in cases where a price increase affects the left most digit, the perception of a price's expensiveness jumps significantly (Thomas 2005). For example, the change in perceived expensiveness between a price change of $2.99 to $3.00 is greater than a price change of $3.00 to $3.01. However, the impact of price changes on perceived expensiveness when moving from the 49-cent to the 50-cent mark was less clear.

Using pricing data gathered for research involving Price Sensitivity Meters, the analysis not only validated the Left-Digit Effect but also demonstrated that moving a price from the 49-cent to 50-cent mark also caused a significant (albeit lesser) increase in perceived expensiveness.

Findings

In the analysis, three scenarios were tested:

• Scenario 1: Price change from 99 cents to the next dollar (e.g. $2.99 to $3.00)
• Scenario 2: Price change from 49 cents to 50 cents (e.g. $2.49 to $2.50)
• Scenario 3: Price change was a penny but did not belong to Group 1 or 2 (e.g. $2.19 to $2.20)

The analysis showed that while the difference in prices for all three cases was a penny, a change in price from 99 cents to the next dollar resulted in a significant increase in expensiveness perceptions. A similar, but smaller effect was also seen when the price changed from 49 cents to 50 cents. Changing the price in all other scenarios had virtually no impact on perceptions around expensiveness.

Table 1: Percentage Changes in Expensiveness ratings

The analysis validated the LDE with significant increases in expensiveness ratings when a change price resulted in the left-most digit to increase. The analysis also demonstrated that when changing a price containing 49 cents to a price containing 50 cents, expensiveness ratings also increase disproportionately higher than a one-cent price increase in other cases.

Methodology

The analysis used the results of pricing research involving the Price Sensitivity Meter (PSM) developed by Dutch economist, Peter van Westendorp (Weiner 2004). The PSM is based on data from four questions (Weiner 2004):

1. At what price would you consider the product to be getting expensive, but you would still consider buying it?
2. At what price would you consider the product too expensive, and you would not consider buying it?
3. At what price would you consider the price to be getting inexpensive and you would consider it a bargain?
4. At what price would you consider the product to be so inexpensive that you would doubt its quality, and would not consider buying it?

The data is aggregated and graphed into curves that represent the incidence of these four price concepts. The points of intersection of the various curves provide researchers and marketers with insight into the acceptable and recommended price ranges.

For this analysis, data was collected across nine separate PSMs and involved a range of products including: fast food sandwiches, beer, juice, medicinal beverages, and insect killer. Price points in each PSM were classified into one of four scenarios:

• Scenario 1: if the price change was from 99 cents to the next dollar (e.g. $2.99 to $3.00)
• Scenario 2: if the price change was from 49 cents to 50 cents (e.g. $2.49 to $2.50)
• Scenario 3: if the price change was a penny but did not belong to Group 1 or 2 (e.g. $2.19 to $2.20)
• Scenario 4: if the price change was more than a penny (e.g. $2.35 to $2.40)

Further, the data was cleaned to exclude price points considered to be outliers. In total, 625 price points were used in the analysis and price points ranged from $0.49 to $10.00. The research involved a total of 4,865 respondents. Respondent profile varied from study to study but generally included both males and females between the ages of 18-64. The majority of the data were collected in Canada.

The impact of the price change was measured in two ways: (1) the simple difference between the percentage of respondents at Price A and Price B and (2) the percentage change between Price A and Price B. The average of the impact was then calculated by the four price groups to illustrate the impact by scenario.

If the LDE existed, it would be expected that the percentage change of moving the price from 99 cents to $1.00 would be greater than a one-cent change between two other price points. Similarly, if the 50-cent effect existed, it would be expected that the percentage change of moving the price from 99 cents to $1.00 would be greater than a one-cent change between two other (non-LDE) price points.

The full results are illustrated in the tables below:

Table 2: Percentage Changes in Expensiveness ratings

Table 3: Differences in Expensiveness ratings

It is interesting to note that while the LDE is apparent when looking at the average % change for the Bargain and Too Cheap ratings, the 50-cent effect is significantly muted. This could suggest that consumers are more sensitive to price increases than price decreases. In fact, there is no change in the "Too Cheap" ratings when the price drops from 50 cents to 49 cents.

Considerations

The analysis demonstrates that perceptions of a price point's expensiveness are dependent on the price point. This does not suggest that marketers should always price at 99-cents or at 49-cents. There are several issues to consider when using the results of this analysis:

Conversion to purchase intent

Research has shown that actual purchase of a product is driven by several factors of which expensiveness is just one. It is important to take the other factors such as the product's relevance and uniqueness in the marketplace into consideration when optimizing pricing. This analysis shows that one-cent price changes can result in vastly different perceptions of expensiveness depending on the ending price-point. The impact on actual purchase is less clear and will vary depending on the other factors.

The analysis can include more industries

While a number of industries have been included in this analysis, the true magnitude is likely different by industry. It is important to keep this limitation in mind when applying the results to industries not involved in the analysis.

Including a wider diversity of price points will likely reveal different effects

While there is a range of price points tested ($0.49 to $10.00), expanding to higher price points will likely reveal different effects. For example, there may be little difference in the impact of a $199 to a $200 price increase when compared against a $199.99 to $200.00 price increase. It is important to keep in mind that the analysis was conducted within a relatively narrow band of price points and to exercise caution when applying these findings to price points outside of this range.

How to price?

It appears from the research that for some industries and when choosing a price within the analyzed range, marketers should use caution when choosing a price point of 99 cents versus $1.00 and 49 cents versus 50 cents. The research suggests that the expensiveness ratings of price points on the dollar and on the 50 cent mark are disproportionately higher than other price points. With expensiveness being an active factor on a consumer's purchase interest, pricing at the $1.00 or 50 cent mark may have a detrimental impact on the purchase interest of the product. For brands struggling to contain their expensiveness perceptions, it is likely better to price at the 99 cent or 49 cent mark.

References

Thomas, Manoj & Morwitz, Vicki (2005). "Penny Wise and Pound Foolish: The Left-Digit Effect in Price Cognition" in Journal of Consumer Research, Vol 32, pages 54-64

Weiner, Jay L & Zacharias, Barry (2004). "Pricing New-to-Market Technologies : An Evaluation of Applied Pricing Research Techniques" in Ipsos White Paper.