Access

You are not currently logged in.

Access your personal account or get JSTOR access through your library or other institution:

login

Log in to your personal account or through your institution.

The Last Name Effect: How Last Name Influences Acquisition Timing

Kurt A. Carlson and Jacqueline M. Conard
Journal of Consumer Research
Vol. 38, No. 2 (August 2011), pp. 300-307
Published by: Oxford University Press
DOI: 10.1086/658470
Stable URL: http://www.jstor.org/stable/10.1086/658470
Page Count: 8
Subjects: Business Marketing & Advertising
Find more content in these subjects: Business Marketing & Advertising
  • Download PDF
  • Add to My Lists
  • Cite this Item

The Last Name Effect: How Last Name Influences Acquisition Timing

Kurt A. Carlson and
Jacqueline M. Conard

In addition to deciding whether to buy an item, consumers can often decide when they buy an item. This article links the speed with which adults acquire items to the first letter of their childhood surname. We find that the later in the alphabet the first letter of one’s childhood surname is, the faster the person acquires items as an adult. We dub this the last name effect, and we propose that it stems from childhood ordering structures that put children with different names in different positions in lines. For example, since those late in the alphabet are typically at the end of lines, they compensate by responding quickly to acquisition opportunities. In addition to responding quicker, we find that those with late alphabet names are more likely to acquire an item when response time is restricted and they find limited time offers more appealing than their early alphabet counterparts.

In 2004, Thomas Zych campaigned as a write-in candidate in the U.S. presidential election on a platform to end the “tyranny” of the alphabetical ordering system. Expanding on the development of this stance, Zych explained that because he grew up with a surname that begins with the letter Z, he knows firsthand how extensive and problematic the adverse effects of the system are: “I spent many years in the back right-hand corner of classrooms, at the ends of lines” (Zych 2004).

Zych’s experience illustrates the extent to which some individuals feel that the use of alphabetical ordering by last name creates an inequitable class system. Specifically, children with surnames that begin with a letter near the beginning of the alphabet enjoy privileged treatment. They are at the beginning of lines, they sit in the front of the class, and they often get first choice when opportunities arise. Those with surnames late in the alphabet face parallel disadvantages. These differential experiences throughout the early formative years of childhood may have implications for behavior throughout one’s life. In the case of Zych, the differential treatment he received was sufficient to compel him to make a rather extreme political statement about the issue.

In psychology, particularly for behaviorism and social learning theory, acquisition timing is the rate at which organisms adapt to their environments. In consumer behavior, acquisition timing refers to the speed with which consumers respond to acquisition opportunities. The speed of response to acquisition opportunities can influence a great number of outcomes that are important for marketers. For example, acquisition timing can influence how quickly consumers respond to direct mail offers, when consumers decide to replace an item that is near the end of its useful life (e.g., refrigerator, roof, car, and shoes), and which consumers become early adopters of a new technology.

This article examines one variable that has a significant and reliable impact on acquisition timing: the last name of the consumer. We propose that this occurs because queuing children based on last names creates differential experience for those with last names near the beginning of the alphabet as compared to those with last names near the end of the alphabet. This inequitable treatment causes those early in the alphabet to be relatively lackadaisical in the face of queuing opportunities where expediency (not last name) determines queue order. At the same time, those late in the alphabet become relatively opportunistic in the face of queuing opportunities. These reactions to queuing opportunities develop into general response tendencies that persist into adulthood. Thus, the speed with which adults respond to acquisition opportunities is influenced by their childhood last name. We dub this proposed effect of childhood last name on acquisition timing the last name effect.

Background and Theoretical Development

Acquisition Timing

While there is a growing body of work on when people consume items they possess (Lowenstein and Prelec 1992), relatively little is known about what dictates the timing of acquisition. What we know is that individuals often put off decisions that involve strong conflict between the choice options (Dhar 1996; Payne, Bettman, and Luce 1996). As such, increases in time pressure (Dhar and Nowlis 1999), concern regarding decision consequences associated with failure (Hogarth, Michaud, and Mery 1980), and similarity in the attractiveness of the choice alternatives (Greenleaf and Lehmann 1995) are associated with delayed choice and thus with delayed acquisition of resources. Such conflicts all stem from the choice context (e.g., the more difficult the trade-offs because of the information format, the more likely individuals are to defer choice), and so they do little to explain variance in acquisition timing across individuals, given a choice setting.

When it comes to individual differences in acquisition timing, one robust finding is that the value an individual ascribes to an object positively influences the speed with which it is acquired (Stanton 1939). Individuals act more quickly to acquire high-value items than they do to acquire low-value items. However, aside from intrinsic valuation effects, we know little about what accounts for between-individual differences in temporal response to opportunities. In this article we propose and test the idea that one’s childhood surname will influence response timing for opportunities to acquire desirable items.

Understanding whether a last name influences acquisition timing is important because customer names are easy to obtain and because there are many decisions in which the decision is not whether to buy but when to buy. Examples include when to renew a favorite magazine subscription, when to buy one’s textbooks, and when to reorder printer supplies. When supplies are limited or when the window of opportunity to purchase a product is finite, heterogeneity in acquisition timing may determine which consumers acquire the product (e.g., before supplies run out or before a sale ends). For example, shopping at a clearance sale, choosing a seat to hear live music, going produce shopping at a farmers’ market—all present consumers with the opportunity to act quickly, which increases the likelihood of securing items of interest, or to act patiently, which increases the chances of missing out on items of interest.

Understanding heterogeneity in acquisition timing is also relevant for successful product diffusion. One of the most used models for product diffusion is the Bass model. This model parses consumers into innovators (who adopt of their own accord) and imitators (who adopt because they imitate someone else who adopted before them). For some products, such as cell phones, imitation is a large determinant of adoption (Sultan, Farley, and Lehmann 1990). For such products, being able to target early adopters is a large determinant of the product’s success (Hoyer and MacInnis 2010). This is because the key to success lies in quickly building a base of product users on which imitation can take root. If it were possible to segment the market into those who were likely to respond quickly versus those who were likely to respond slowly to an acquisition opportunity, marketers could use this segmentation scheme to build such a customer base at a lower cost. Of course, the key is to find a good, observable proxy for acquisition timing.

The Last Name Effect

Previous research has revealed various influences that one’s name can have on one’s choices. For example, Nuttin (1985) found that the first letter of one’s surname can influence preferences for items with a similar spelling, a result he dubbed the name-letter effect. The name-letter effect has been shown for one’s choice of spouse (Jones et al. 2004) and products (Brendl et al. 2005), as well as one’s academic achievement (Nelson and Simmons 2007). It seems that people gravitate to things that they see as reflections of themselves (including things that share properties with their surname) because they have positive beliefs about themselves.

Although people tend to like their names and things that are associated with their names, society does not treat all names as equal. Many organizations (e.g., elementary and middle schools) rely on alphabetical order of surnames to determine the order in which children access various opportunities. We propose that the inequitable treatment imposed by such structures gives rise to response tendencies that persist into adulthood.

The idea that children who are routinely forced to wait would develop a tendency toward expediency is consistent with research on the effect of goal interference on goal pursuit. This research finds that when a goal is interfered with, the goal increases in activation, causing people to pursue the goal with greater speed and effort once the interference is removed (Chartrand et al. 2008). Additionally, goals that are routinely pursued by consumers in certain contexts can be automatically activated by those contexts (Bargh et al. 2001). Thus, it is possible that repeated delays imposed on children whose last names are late in the alphabet create in these individuals a chronic expediency motive that is automatically activated by an opportunity to determine their position in a queue. The repeated pursuit of such a motive can produce a pattern of behavior that we refer to as a response tendency.

Our contention is that children develop temporal response tendencies to manage the inequities that arise from the use of last names to order their environment. If these response tendencies persist into adulthood, then there should be an association between childhood last name and the speed with which adults acquire items in settings where the individual has control over acquisition timing. Our research hypothesis is that adults with childhood surnames beginning with letters late in the alphabet will respond relatively quickly to opportunities to acquire items, while adults with childhood surnames beginning with letters early in the alphabet will respond relatively slowly to the same acquisition opportunities. Specifically, we expect to find a negative association between response time to acquisition opportunities and the depth into the alphabet of the first letter of the responder’s childhood surname.

Study 1

Study 1 was designed to test for the last name effect by examining how quickly individuals respond to an offer to acquire free basketball tickets. We expect individuals with surnames beginning with letters late in the alphabet will respond more quickly to the opportunity, relative to individuals with surnames beginning with letters early in the alphabet.

Methods

MBA students (most of whom are typically unmarried) were given the opportunity to receive up to four free tickets to attend a top-ranked women’s basketball game. The announcement of the free tickets was made by e-mail to the university student e-mail accounts via an e-mail distribution list. Students interested in receiving up to four free tickets only had to respond by e-mail before all the free tickets were spoken for. The limited supply of tickets was allocated according to the order in which the requests arrived. The sample of 76 respondents consists of those students who replied before all the tickets were spoken for—at which time another e-mail was sent to the student body indicating that all the tickets had been allocated. Response times were measured by computing the difference between the time the offer was sent out by e-mail and when each reply e-mail was received (in minutes).

The first letter of each respondent’s last name was derived from their e-mail address. This was possible because each student at the school at which the study was run had an e-mail address consisting of his or her initials and a number. Since aliases were not permitted when this study was run, the letter immediately preceding the first number in the student’s e-mail address was the first letter of the student’s surname. For analysis, the first letter of each respondent’s last name was transformed into a number equivalent (i.e., A = 1; Z = 26).

Results

The average response time to the offer was 22.7 minutes. More importantly, there was a reliable negative correlation between response time and the number equivalent of the last name (r = −.271, p = .018). This last name effect can also be seen by comparing average response time for those with a last name beginning with one of the last nine letters of the alphabet (M = 19.38 minutes, SD = 7.12) to that of those with a last name beginning with one of the first nine letters of the alphabet (M = 25.08 minutes, SD = 10.41; t(48) = 2.24, p = .030, d = .647). In sum, these results revealed a last name effect for individuals faced with an opportunity to obtain items (basketball tickets) from a constrained supply.

We looked for evidence of model misspecification, but we found none. The last name effect appeared whether or not time was transformed by the natural log, and a White test suggested that the residuals were homoskedastically distributed (F(2, 73) = 2.20, p > .10). Thus, there was no evidence of heteroskedasticity in the residuals of the simple linear model. Moreover a test of variance in response times for those early versus late in the alphabet revealed insufficient evidence to reject the null of equal variances using Levene’s test (p > .25). Correlation of the absolute value of the residuals with last name revealed no association, suggesting that the effect was relatively continuous throughout the alphabet. We also looked for evidence that there might be some outliers contributing excessive influence on the observed correlation. Using standard cutoff values for high leverage observations, we found no observations that had the potential to exert high leverage on the observed relationship. Next, we tested for model misspecification by conducting a pure error lack-of-fit test on the regression of response time on last name. The analysis revealed no evidence of model misspecification (F(18, 56) = .68, p = .81).

Finally, since last names are not uniformly distributed throughout the alphabet, and since the number equivalent of the first letter of each respondent’s surname is a uniform metric, we examined whether using cumulative frequency of last names in the United States instead of the number equivalent produced different results. We computed the cumulative frequency of surnames for each letter of the alphabet from 1990 Census data for the 60,000 most common surnames in the United States (e.g., 3.4% of surnames in the United States begin with the letter A, 12.5% begin with either A or B, and 20.3% begin with either A, B, or C). Using this metric to translate each participant’s last name into a new independent variable, we again found evidence of the last name effect. That is, there was a significant negative correlation between response time and depth into the alphabet as measured by cumulative frequency of surname first letter (r = −.16.6, p = .037).

Discussion

This study showed that consumers’ temporal response to an acquisition opportunity differed depending on the first letter of their last name. Consumers with names starting near the end of the alphabet responded more quickly than those with last names starting at the beginning of the alphabet. This occurred for a desirable item that was limited in supply (i.e., free basketball tickets). The next study attempts to replicate the effect for a desirable but unconstrained item, while tracing the roots of the effect to the childhood name.

Study 2

Given the novelty of the last name effect, it seemed prudent to test it in a different setting. To this end, we used a traditional experimental context in which members of an on-line panel were invited to participate in a study in exchange for compensation. This sampling frame should produce a sample of older adults, many of whom are married and some of whom have changed their last name. This allows us to examine if the last name effect is tied to one’s childhood name or to one’s current name. If the effect originates in childhood, then for adults who changed their surname, the last name effect should exist for their childhood surname but not for their adult surname. We test this hypothesis in the current study by examining response times to participate in an on-line survey as a function of childhood surnames for those who have not changed their name and adult and childhood surnames for those who have changed their name.

This study also extends the exploration of the last name effect in another way. The incentive for completing the online survey was entry into a drawing to win $500, with a 1-in-500 chance of winning. Thus, respondents expected that there was only a small chance (.2%) that they would obtain the desired item. (In reality, 280 individuals responded to the survey, so the true odds were 1-in-280.)

Methods

Participants were 280 adults (average age = 39.1 years; 51.2% female) who responded within 1 day to an e-mail invitation to participate in a survey in exchange for an opportunity to win $500. Participants’ response times were calculated from the time the e-mail invitation was sent to the time each participant logged into the platform to take the study. The first letter of each respondent’s adult last name and the first letter of the last name during childhood were obtained at the end of a 5-minute survey that was unrelated to this study.

Results

The average response time (from e-mail delivery to log in) for those who completed the survey was 389.5 minutes. For the sample as a whole, there was a significant negative correlation between childhood surname and response time (r = −.128, p = .033) but an insignificant negative correlation between first letter of adult surname and response time (r = −.070, p = .241). This difference in correlations stemmed from individuals who had changed their name (n = 101). When just the subset of name changers was examined, no correlation between adult surname and response time was observed (r = .000, p = .999). However, there was a negative correlation between response time and childhood surname for these name changers (r = −.154, p = .122).

As in study 1, we found no evidence of model misspecification. Specifically, the residuals were distributed appropriately whether time to respond was log transformed or not, the variance in response times did not differ across those early versus late in the alphabet, and the correlation of the absolute value of the residuals with last name revealed no association. A White test for heteroskedasticiy (F(2, 276) = 2.27, p > .10) suggested that the residuals were distributed uniformly. Moreover, we found no evidence of high leverage outliers, and the pure error lack-of-fit test was again insignificant (F(20, 257) = 1.15, p = .302). Like the data in study 1, these analyses suggest that the effect was continuous throughout the alphabet, a point that is consistent with those early in the alphabet responding relatively slowly and those late in the alphabet responding relatively quickly.

Given that the majority of name changers in our sample were women (98.0%), the possibility exists that the null result for name changers was driven by a gender effect, with women exhibiting a slower response on average than men. Analysis does not support this possibility. The average response time of women in our sample (M = 377 minutes) did not differ from the average response time for men (M = 402 minutes; t(278) = .52, p = .60). A multivariate regression of response time on gender and last name revealed a significant effect of last name (t(278) = −2.13, p < .05) but no effect of gender (t(278) = −.49, p = .63). Likewise, a multivariate model including an interaction term for gender and last name yielded no main effect of gender or interaction with last name on response time (both p > .85).

Discussion

These data support the claim that the last name effect emerges from a childhood response tendency. For those who changed their name, the effect was observed for childhood surname, not adult surname. These data also suggest that gender plays no direct role in the last name effect. Thus far we have examined only the data of those who responded to our offers. In addition to examining the response times of responders, the next study also examines the last names of nonresponders.

Study 3

Although studies 1 and 2 established the last name effect, sourced it to childhood surnames, and demonstrated that it generalizes to both young and old adults, we did not gather information on nonresponders in these studies. Thus, it was not possible to compare the typical last name of those who responded to that of those who did not. The last name effect makes a clear prediction in this regard: given a limited time in which to respond, the response sample should (on average) consist of more people whose last names are nearer to the end of the alphabet than the nonresponse sample. We could not examine this issue in study 2 because we did not possess last name information for the entire sampling frame, which was an on-line national panel of more than 1 million people. In theory we could have constructed the full sampling frame for study 1. However, since the full frame consisted of more than 1,200 MBA students, the nonresponder sample would have dwarfed the responder sample (n = 76), impeding our ability to find a difference in last names between the samples. In study 3, we knew the last names of all people in the sampling fame, and the nonresponder and responder samples were of roughly equal size. This allowed us to test if the responder sample consisted on average of more individuals near the end of the alphabet.

Studies 1 and 2 used time between when the e-mail invitation was sent and when the e-mail response was received as the dependent variable. This raises the possibility that the last name effect is driven partially by how often those late in the alphabet check their e-mail. Although frequent checking of one’s e-mail could be one result of an expedient response tendency, it is not our focus in this article. Therefore, the current study explores a situation where everyone in the sampling frame received the offer at the same time.

Methods

Undergraduate seniors and graduate students of legal drinking age who were enrolled in a wine evaluation class (consisting of 302 students) were given the opportunity to participate in a 45-minute wine choice study in exchange for $5 and a bottle of their selected wine. The study was announced at the end of class to ensure that all students had an equal opportunity to respond (via e-mail) upon leaving class. Because the announcement was made at the end of class, which was followed by a uniform break during which students could respond to e-mail, there was no known delay differentially affecting any group in the study. Individuals who may have had something else to do during the break would show up as random noise across the alphabet.

In total, 160 students (average age = 22.2 years, 49.4% female) responded in the first 24 hours. These individuals constitute the responder sample for the analysis below. As in study 1, response times (measured in minutes from when the class was adjourned) constitute the dependent variable. The independent variable was the number equivalent (1 to 26) of the first letter of each respondent’s last name, which was discerned from responders’ e-mail addresses. The e-mail addresses were as described in study 1, so the letter immediately before the first number in the address was the first letter of each responder’s last name.

Results

Average response time was 359.8 minutes—just under 6 hours. There was a significant negative correlation between last name and response time (r = −.156, p = .045). As in study 1, the last name effect was also evidenced by shorter response times for participants with last names beginning with one of the last nine letters of the alphabet (M = 329.1 minutes, SD = 123.6), relative to response times for those with a last name beginning with one of the first nine letters of the alphabet (M = 393.8, SD = 163.8; t(111) = 2.34, p = .021, d = .444). In other words, those with last names beginning with letters near the end of the alphabet responded roughly 1 hour faster than those with names near the front of the alphabet. Again, we looked for evidence of model misspecification (using the same tests as in studies 1 and 2), but we found no such evidence.

We also tested for gender effects as both a main effect on acquisition timing and as an interaction term in a multivariate regression of response time on last name, gender, and the interaction term. In all models, whether gender was examined in isolation, in conjunction with last name, or in conjunction with last name and the interaction between these two variables, gender and its interaction with last name played no role in determining response time (all p > .90).

Finally, we examined the last names of those who did not respond. We built the last name letter list for nonresponders as follows. We extracted the first letter of each last name on the course roster and removed from this the last name letters from the responder database. For example, if there were 22 A’s on the class roster (22 students with a last name beginning with the letter A), and 10 A’s in the responder database, the nonresponder last name list would have 12 A’s on it. We do not have age or gender for nonresponders, but based on our reading of the first names on the class roster, the class was roughly 50% female. Since the responder sample was 49.4% female, the nonresponder sample was roughly 50% female. Comparing the average position of the first letter of the last name for nonresponders (M = 9.88) to the average letter of responders (M = 12.29) revealed the expected result. Responders (on average) had last names with letters later in the alphabet than nonresponders (t(300) = 2.81, p < .01). Put differently, people with last names later in the alphabet were more likely to sign up to participate in the study.

Discussion

These data provide additional support for the last name effect. The effect existed for individuals who knew they would have to incur a cost to acquire the $5 and the bottle of wine (i.e., they would have to give up 45 minutes of their time). This was true even though respondents were not led to believe that the items were scarce. This last point is noteworthy as many marketing offers are time constrained but not quantity constrained. For example, it is common practice to provide a rain check if a promoted product is out of stock during a promotion cycle. As in study 2, gender was unrelated to acquisition timing, and it did not interact with the effect of last name on acquisition timing.

The current study also found that the last name effect can produce a sample selection phenomenon in which those early in the alphabet are more likely to be nonresponders. For marketers, this raises the important question of whether those late in the alphabet will be more likely to respond to a limited time offer. In the next study we examine a situation in which individuals must decide if they would be willing to undertake effort to secure a limited time promotional offer. If the last name effect extends to this setting, then those with last names late in the alphabet will be more likely to pursue the offer. Before introducing the next study, we want to summarize the findings so far and to examine the issue of continuity of the effect more directly.

In studies 1–3, there was a negative correlation between last name and response time. Using correlation to test this relationship assumes that the relationship is linear, an assumption that could not be rejected on the basis of model specification tests. Nevertheless, to better understand if the last name effect is driven jointly by slower responses from those early in the alphabet and quicker responses at the end of the alphabet, we wanted to examine this question of linearity more directly. To do so we analyzed the data from the studies above by transforming each participant’s response time into a z-score, using the mean and standard deviation from the sample to which the participant belonged. We combined the data for studies 1–3 and correlated z-scored response time with last name code. This revealed a strong last name effect (r = −.158, p < .001). Next, we plotted average z-scored response times for the four quartiles of the last name equivalent variable (see fig. 1). This figure reveals that those with surnames early in the alphabet (first quartile) responded slowest and that each subsequent quartile responded faster. This monotonicity indicates that the last name effect is driven both by those early and those late in the alphabet, while the linearity of this plot suggests that the relationship between last name and acquisition timing is uniform throughout the alphabet.

Figure 1. 
Mean Z-Scored Response Time for Studies 1–3 as a Function of Last Name Quartile

Study 4

As discussed above, the last name effect should influence acquisitions when the window of opportunity is fixed. This is because consumers with surnames that begin with letters late in the alphabet will tend to act quickly and so will be more likely to take advantage of limited-time offers. Since rushing to take advantage of such offers is a natural part of how these consumers act, they might even find such promotions to be more appealing than would individuals with surnames that begin with letters early in the alphabet, who have a more lackadaisical response. This study tests this idea by examining the influence of last name on the appeal and purchase intent of a product that was discounted by 20% but only while supplies lasted.

Methods

Participants were 41 undergraduate students who were financially compensated for participating in this and other unrelated studies. Each participant read a choice scenario and answered three questions. The scenario asked participants to assume that they needed a new backpack and that upon passing the bookstore they noticed a poster saying that the bookstore was running a 20% off sale on a name brand backpack. The scenario said that the promotion was good “while supplies last.” In the scenario participants were told to assume that they did not currently have their wallet with them, so buying at this instant was not an option and that it would take about 15 minutes (round trip) to go home and get their wallet. After reading the vignette, participants answered three questions. The first was a question about the appeal of the discount: “How appealing is the discount?” Responses were captured on a 9-point semantic differential scale (1 = not at all appealing; 9 = extremely appealing). The second question was a purchase intent question: “How likely are you to go home to get your wallet so you can buy the backpack today?” Responses were collected on a 9-point scale (1= very unlikely; 9 = extremely likely). Finally, participants were asked to write down the first letter of their last name. Based on the results above, we expect responses to the purchase question, which probes the likelihood participants will act today to take advantage of the promotional offer, to be positively correlated with the depth of participants’ last name in the alphabet. Moreover, since those with last names late in the alphabet are accustomed to acting quickly, limited offers like the one in this vignette, which require expedient action to be realized, should be more congruent with their response tendency and so may be more appealing to them.

Results

To test for the last name effect, we correlated last name (coded as in previous studies) with response to the appeal and purchase intent questions. The correlation between last name and purchase intent was positive (r = .267, p < .05), indicating that the last name effect extended to a marketing promotion context. That is, those late in the alphabet indicated that they were more likely to act on the promotion than were those early in the alphabet. The correlation between last name and appeal of the promotional offer was also positive (r = .280, p < .05), indicating that the offer’s appeal increased with depth in the alphabet of the first letter of the participant’s surname.

Discussion

This study found that the last name effect generalized to common consumer purchase settings. In particular, those with surnames beginning with letters late in the alphabet indicated that they would be more likely to act to take advantage of a 20% off “while supplies last” promotion than those with surnames beginning with letters earlier in the alphabet. If supplies were truly constrained, those with surnames deeper into the alphabet would presumably have accounted for a disproportionate share of sales. Although the scenario was designed to fit the life of a college student, its characteristics (e.g., awareness of a limited time sale where one must travel to take advantage of the sale before time or supplies expire) are common in everyday, real-life consumer markets.

General Discussion

The studies above reveal a robust effect of the first letter of one’s childhood surname on temporal responses to acquisition opportunities. The deeper into the alphabet the first letter of one’s last name, the more quickly (on average) the individual responded to an opportunity to acquire an item. This last name effect occurred when the average response time was in minutes and in hours, in situations where the opportunity was constrained loosely or tightly, and when the items were real (basketball tickets, cash, and wine) or hypothetical (limited time price promotion on a backpack). The effect occurred only for childhood surnames (not for adult surnames of those who changed their name), indicating that it is driven by response tendencies adopted during childhood.

Response times were slowest for those with last names beginning with letters at the front of the alphabet, and response times were fastest for those with last names near the end of the alphabet. This suggests that those at both ends of the alphabet respond to the use of last name as an organizing structure. Those early in the alphabet typically develop a lackadaisical response tendency, while those late in the alphabet develop a relatively expedient response tendency. A question for future research is what intermediate purpose these tendencies serve. An expedient tendency helps consumers acquire items while it also helps them avoid missing out on items. Thus, acting quickly is both a means to acquisition and insurance against acquisition failure. Future research might examine if those who act quickly get more utility from acquisition than those who act slowly, if they experience greater disutility from a failed acquisition, or both. A tendency to act slowly could help those early in the alphabet feel like they are laid back. If so, then those with last names beginning with letters near the front of the alphabet should see themselves as more laid back and they should derive greater utility from being laid back than those with last names near the end of the alphabet. Whatever the answer to these questions, the finding that one’s childhood last name has such a consistent effect on adult behavior speaks to the profound influence that alphabet-based organizing structures can have on how consumers respond to opportunities.

Implications

The ease of obtaining information about one’s last name, the ubiquity of ordering structures that rely on last names, and the reality that nearly everyone has a last name (some pop stars aside) means that the last name effect is particularly relevant for marketers. For example, marketers might use last names to create a priori segmentation schemes that leverage the last name effect. Such segmentation schemes could partition prospective customers into those who are likely to respond quickly and those who are likely to respond more slowly to various acquisition opportunities. Consumers with surnames deeper in the alphabet should respond faster to direct mail opportunities, making them an ideal target for promotions where supplies are truly limited. Likewise, if database marketers want to assess the effectiveness of a promotional offer in the shortest possible time, then oversampling individuals with surnames late in the alphabet is one way to shrink the testing horizon. Finally, if marketers want to quickly build a customer base on which imitation effects can take root, they might focus on those who have names late in the alphabet.

As noted above, consumers who delay purchasing a noncommodity risk missing the acquisition opportunity altogether. Our findings suggest that those with childhood surnames that begin with letters early in the alphabet are more likely to behave in ways that expose them to such risk. It is possible that repeated exposure to such risks gives rise to strategies for coping with this risk. For example, those with childhood surnames that begin early in the alphabet might be more willing to accept a substitute for a missed opportunity, they might be more likely to exclaim sour grapes when an opportunity has been lost, and they might be more likely to believe that the market does not work as efficiently as it should. Future research might explore such questions.

Finally, the ability to generalize research findings from a sample to a population of interest requires that the sample be representative of the population. Our results suggest that many studies suffer from a sample-selection bias in which quick-responding (i.e., late alphabet) individuals are overrepresented. This bias is most worrisome when the research question relates to the timing of the behavior or the fervency with which individuals exhibit the behavior of interest. One example is a research study designed to gather purchase intent for a product with a late-alphabet name (e.g., Tide with Bleach). Such a study may overstate actual demand because late name people are overrepresented in the sample and because people tend to like brands that begin with letters that they have in their name (Brendl et al. 2005).

Conclusion

The studies of this article established the last name effect, a tendency for those with surnames beginning with letters early in the alphabet to respond to acquisition opportunities relatively slowly while those with surnames beginning with letters late in the alphabet respond relatively quickly. This effect occurred in a variety of settings, with diverse acquisition items and samples. The last name effect contributes to a growing body of work that uses a life perspective in understanding consumer behavior by exploring the role that childhood organizational systems have on adult decision making. In doing so, we have identified a ubiquitous factor (childhood last name) that has a consistent influence on acquisition timing.

References

  1. Bargh, John A., Peter M. Gollwitzer, Annette Lee-Chai, Kimberly Barndollar, and Roman Trötschel (2001), “The Automated Will: Nonconscious Activation and Pursuit of Behavioral Goals,” Journal of Personality and Social Psychology, 81 (December), 1014–27.
  2. Brendl, C. Miguel, Amitava Chattopadhyay, Brett W. Pelham, and Mauricio Carvallo (2005), “Name Letter Branding: Valence Transfers When Product-Specific Needs Are Active,” Journal of Consumer Research, 32 (December), 405–15.
  3. Chartrand, Tanya L., Joel Huber, Baba Shiv, and Robin Tanner (2008), “Nonconscious Goals and Consumer Choice,” Journal of Consumer Research, 35 (August), 189–201.
  4. Dhar, Ravi (1996), “The Effect of Decision Strategy on Deciding to Defer Choice,” Journal of Behavioral Decision Making, 9 (December), 265–81.
  5. Dhar, Ravi and Stephen M. Nowlis (1999), “The Effect of Time Pressure on Consumer Choice Deferral,” Journal of Consumer Research, 25 (March), 369–82.
  6. Greenleaf, Eric A. and Donald Lehmann (1995), “Reasons for Substantial Delay in Consumer Decision Making,” Journal of Consumer Research, 22 (2), 186–99.
  7. Hogarth, Robin M., Claude Michaud, and Jean-Louis Mery (1980), “Decision Behavior in Urban Development: A Methodological Approach and Substantive Considerations,” Acta Psychologica, 45 (August), 95–117.
  8. Hoyer, Wayne D. and Deborah J. MacInnis (2010), Consumer Behavior, Mason, OH: South-Western College.
  9. Jones, John T., Brett W. Pelham, Maricio Carvallo, and Mathew C. Mirenberg (2004), “How Do I Love Thee? Let Me Count the Js: Implicit Egotism and Interpersonal Attraction,” Journal of Personality and Social Psychology, 87 (5), 665–83.
  10. Lowenstein, George and Drazen Prelec (1992), “Anomalies in Intertemporal Choice: Evidence and Interpretation,” Quarterly Journal of Economics, 107 (2), 573–97.
  11. Nelson, Leif D. and Joseph P. Simmons (2007), “Moniker Maladies: When Names Sabotage Success,” Psychological Science, 18 (12), 1106–12.
  12. Nuttin, Jozef M., Jr. (1985), “Narcissism beyond Gestalt and Awareness: The Name Letter Effect,” European Journal of Social Psychology, 15 (3), 353–61.
  13. Payne, John W., James R. Bettman, and Mary Francis Luce (1996), “When Time Is Money: Decision Behavior under Opportunity-Cost Time Pressure,” Organizational Behavior and Human Decision Processes, 66 (2), 131–52.
  14. Stanton, Frank (1939), “Notes on the Validity of Mail Questionnaire Returns,” Journal of Applied Psychology, 23 (February), 95–104.
  15. Sultan, Fareena, John U. Farley, and Donald R. Lehmann (1990), “A Meta-analysis of Applications of Diffusion Models,” Journal of Marketing Research, 27 (1), 70–77.
  16. Zych, Thomas (2004), Web site, http://www.zych4prez.com.

References

  1. Bargh, John A., Peter M. Gollwitzer, Annette Lee-Chai, Kimberly Barndollar, and Roman Trötschel (2001), “The Automated Will: Nonconscious Activation and Pursuit of Behavioral Goals,” Journal of Personality and Social Psychology, 81 (December), 1014–27.
  2. Brendl, C. Miguel, Amitava Chattopadhyay, Brett W. Pelham, and Mauricio Carvallo (2005), “Name Letter Branding: Valence Transfers When Product-Specific Needs Are Active,” Journal of Consumer Research, 32 (December), 405–15.
  3. Chartrand, Tanya L., Joel Huber, Baba Shiv, and Robin Tanner (2008), “Nonconscious Goals and Consumer Choice,” Journal of Consumer Research, 35 (August), 189–201.
  4. Dhar, Ravi (1996), “The Effect of Decision Strategy on Deciding to Defer Choice,” Journal of Behavioral Decision Making, 9 (December), 265–81.
  5. Dhar, Ravi and Stephen M. Nowlis (1999), “The Effect of Time Pressure on Consumer Choice Deferral,” Journal of Consumer Research, 25 (March), 369–82.
  6. Greenleaf, Eric A. and Donald Lehmann (1995), “Reasons for Substantial Delay in Consumer Decision Making,” Journal of Consumer Research, 22 (2), 186–99.
  7. Hogarth, Robin M., Claude Michaud, and Jean-Louis Mery (1980), “Decision Behavior in Urban Development: A Methodological Approach and Substantive Considerations,” Acta Psychologica, 45 (August), 95–117.
  8. Hoyer, Wayne D. and Deborah J. MacInnis (2010), Consumer Behavior, Mason, OH: South-Western College.
  9. Jones, John T., Brett W. Pelham, Maricio Carvallo, and Mathew C. Mirenberg (2004), “How Do I Love Thee? Let Me Count the Js: Implicit Egotism and Interpersonal Attraction,” Journal of Personality and Social Psychology, 87 (5), 665–83.
  10. Lowenstein, George and Drazen Prelec (1992), “Anomalies in Intertemporal Choice: Evidence and Interpretation,” Quarterly Journal of Economics, 107 (2), 573–97.
  11. Nelson, Leif D. and Joseph P. Simmons (2007), “Moniker Maladies: When Names Sabotage Success,” Psychological Science, 18 (12), 1106–12.
  12. Nuttin, Jozef M., Jr. (1985), “Narcissism beyond Gestalt and Awareness: The Name Letter Effect,” European Journal of Social Psychology, 15 (3), 353–61.
  13. Payne, John W., James R. Bettman, and Mary Francis Luce (1996), “When Time Is Money: Decision Behavior under Opportunity-Cost Time Pressure,” Organizational Behavior and Human Decision Processes, 66 (2), 131–52.
  14. Stanton, Frank (1939), “Notes on the Validity of Mail Questionnaire Returns,” Journal of Applied Psychology, 23 (February), 95–104.
  15. Sultan, Fareena, John U. Farley, and Donald R. Lehmann (1990), “A Meta-analysis of Applications of Diffusion Models,” Journal of Marketing Research, 27 (1), 70–77.
  16. Zych, Thomas (2004), Web site, http://www.zych4prez.com.