August 27, 2007
No, there's not a typo in the title.
Back in late March, 2007, TrueDelta posted a blog entry on corruption in dealer sales satisfaction surveys. All auto manufacturers survey customers about their sales experience, and in many cases the dealer and salesperson only receive a large bonus if the customer gives them all perfect scores. In response, car salespeople often ask or even beg for perfect scores. Some go so far as to offer gifts--an oil change, a tank of gas, a restaurant gift certificate--if the customer will fill the survey out in the sales manager's office. In at least one case, a customer who gave the dealer poor scores was subsequently told they were no longer wanted as a customer. (These surveys are usually not anonymous, so dealers learn which customers were naughty and which were nice.)
But all of this was anecdotal hearsay. How common is such corruption of the new car sales satisfaction survey process, really? To find out, in May TrueDelta asked U.S. and Canadian panel members who bought a 2003 or newer car as a new car to complete a brief survey concerning automobile manufacturers' sales satisfaction surveys.
This report on the results of that survey on surveys includes the following sections:
- Methods and Sample
- Overall Results
- Effects of Gender and Age
- Results by Brand
- About TrueDelta
- Appendix: Respondents' Comments
Methods and Sample
About 7,000 people were asked to complete a survey. They provided 1,728 responses, with few people responding for more than one car.
The response rate of about 23 percent is lower than with TrueDelta's main survey, the Vehicle Reliability Survey, for a number of reasons. First of all, many of those asked to complete the survey were not eligible because they did not buy the vehicle new or did not personally purchase the vehicle. In other cases, chiefly with the older model years, their memory was not clear enough to respond (those who could not remember were told not to respond). Respondents indicated the year of the purchase. Only two percent, nine percent, and 11 percent occurred in 2002, 2003, and 2004, respectively. Response rates were much higher for more recent purchases. Finally, because the results remained steady as responses came in, the usual follow-up emails were not sent.
Respondents indicated their age and gender. Only seven percent were female. This is probably because TrueDelta surveyed members of its Vehicle Reliability Research Panel, who tend to be the person in charge of maintaining the household's vehicles. Respondents were more evenly distributed in terms of age:
- Under 20: 1 percent
- 20s: 16 percent
- 30s: 26 percent
- 40s: 20 percent
- 50s: 21 percent
- 60s: 11 percent
- 70+: 2 percent
The sample was not random. A random sample would have been ideal, especially for those questions that have a subjective aspect. That said, no bias is evident in the results. It likely helped that
the survey was not open to the general public or publicized beyond the panel, which was recruited for a different purpose.
In general, people who've had negative experiences tend to be more likely to respond to such a survey, such that negative experiences will seem more common than they actually are. But the worst dealer practices were reported by few respondents. In general, while the mildest forms of corruption seem common, the relatively extreme forms seem rare, and much rarer than anecdotal evidence suggests.
In the following subsections, responses to the Dealer Sales Experience Survey Survey are discussed question by question. The questions are reproduced verbatim in the section headings.
What scores did you give the dealer on the manufacturer's sales experience survey?
Sixteen percent of respondents did not receive a survey, and another five percent did not complete it. While no doubt the response rate to such surveys
is nowhere near 95 percent, it is likely that those who did not respond to the sales experience survey also did not respond to TrueDelta's survey concerning
such surveys. Some of those who did respond here commented that they did not respond to the manufacturers' survey because they did not want to give perfect scores, and felt they would have had to if they had responded.
Those who did complete the manufacturers' surveys rated their dealers as follows:
- below average scores: 5 percent
- average scores: 8 percent
- above average scores, but not all perfect scores: 52 percent
- all perfect scores: 36 percent
Clear evidence of grade inflation, since 88 percent of dealers were rated above average. But are dealers' attempts to corrupt the process a major factor?
When you purchased your car, what did the salesperson or manager say about the survey?
- they never mentioned a survey: 25 percent
- they mentioned that I would receive a survey (but none of the following): 30 percent
- they asked me to let them address any problems rather than reporting them: 36 percent
- they asked me to give them perfect scores, but did not beg: 28 percent
- they begged me for perfect scores: 8 percent
- they said they would only get a bonus if I gave them perfect scores: 9 percent
- they asked me to bring the survey to the dealer and fill it out while they watched: 2 percent
- they asked me to bring the blank survey to the dealer so they could fill it out: 2 percent
- they offered me a gift in exchange for one of the above: 2 percent
Here we see that about 75 percent of dealer personnel mentioned the survey to car buyers, and that about 45 percent tried to influence survey responses.
(The numbers will not sum to 100 because multiple responses were allowed.)
In most cases, though, the tactics used were relatively mild, asking for perfect scores or for a chance to correct any problems. The latter, to the extent that the dealers actually addressed the problems, might even be seen as a positive effect. Because of the surveys, dealers might be more willing to address problems than they would otherwise be.
In contrast, the relatively extreme forms of corruption were much less common than anecdotal evidence suggested. They were each reported by just two percent of respondents.
How much did the dealer's statements regarding the survey influence your responses?
- no pressure - doesn't apply: 26 percent
- no effect: 41 percent
- did not respond as a result: < 1 percent
- gave them much lower scores as a result: < 1 percent
- gave them slightly lower scores as a result: 3 percent
- gave them slightly higher scores as a result: 10 percent
- gave them much higher scores as a result: 2 percent
Note: the last four percentages only include those who completed a survey. They would have been a bit lower otherwise. Even with this adjustment, they're fairly low. About one in eight car buyers reported inflating their scores in response to dealer pressure. This likely understates the actual amount of the influence, as people generally do not like to admit to having been influenced, even to themselves. Courses on consumer behavior teach that, especially with mild forms of persuasion, that if you can influence someone to buy a product, they will believe they bought it because they wanted to buy it, and not because of the influence. A similar effect is probably present here, given the widespread mild attempts to influence scores and the large percentage of high scores.
How much pressure did you feel to give them perfect scores?
- no survey mentioned - doesn't apply: 25 percent
- no pressure: 48 percent
- a little pressure: 22 percent
- a lot of pressure: 5 percent
About one in four respondents reported feeling pressured to give the dealer perfect scores. Combine this with the previous question, and about half of those
who reported feeling pressured also reported giving the dealer higher scores as a result.
How did the dealership react to your survey responses?
- no reaction: 75 percent
- they thanked me: 24 percent
- they gave me a gift: 7 percent
- they offered to address problems I reported: 4 percent
- they actually did address problems I reported: 3 percent
- they complained about my response: 2 percent
- they were rude with me: 1 percent
- they were angry with me: 1 percent
- they told me I was no longer wanted as a customer: < 1 percent
These surveys are usually not anonymous. The dealer learns who responded, and what their responses were. Most dealer reactions were positive: about a quarter of the time
the car buyer was thanked for responding. Seven percent of the time the car buyer received a gift. Based on comments, this gift was most often a free oil change. Given the earlier response,
most of the time the car buyer was not told they'd receive this gift before they completed the survey. Negative dealer reactions were much rarer than anecdotal evidence suggested they would be.
Effects of Gender and Age
In nearly all cases, the results do not differ by more than a few points between men and women. The main difference: women were much more likely to feel pressured. Thirty-seven percent reported
feeling at least a little pressure, compared to 27 percent of men. This does not appear to have significantly influenced scores on the surveys, though. With a larger sample of female car buyers
some additional influence might become evident, but it is not likely to be large based on the current results.
Similarly, results do not vary much by age. With buyers between the ages of 20 and 70, no clear patterns are evident. The largest difference comparing an age group with the mean is six points, with nearly all differences two or fewer points.
Larger differences might be present with buyers over 70, most notably in the percentage giving the dealer perfect scores (55 percent versus 40 percent for buyers in their 60s, and 36 percept overall), but the sample size of such buyers is too small to be conclusive.
Many people have suggested that age is a factor in satisfaction surveys, with older people allegedly less likely to be critical. Based on this survey, age does not appear to be a significant factor until 70 or so.
Results by Brand
Brands with at least 40 responses were separately analyzed. Because most of the sample sizes at the brand level are small, only
especially large differences are noted below.
Acura (87 responses)
More than minimal attempts at influence rare, even non-existent.
Dealers appear to care less about these surveys than the average dealer, and were only about half as likely to react to a buyer's survey responses.
BMW (46 responses)
Sixty-three percent of BMW salespeople asked or even begged for perfect scores, compared to the average of 36 percent.
Similarly, 57 percent asked buyers to let them address any problems rather than reporting them, compared to the average of 36 percent.
Not surprisingly, BMW buyers were more likely to report feeling pressured to provide high scores (44 vs. 27 percent) and giving dealers higher scores than they otherwise would have as a result (21 vs. 12 percent).
These tactics appear to be effective: 67 percent of BMW buyers reported giving the dealer all perfect scores, compared to the average of 36 percent. No BMW buyers reported giving the dealer average or lower scores.
Chevrolet (96 responses)
Chevrolet buyers were more likely than the average buyer to report giving dealers perfect scores (46 vs. 36 percent). Everything else close to the averages.
Chrysler (42 responses)
Chrysler dealers were less likely to attempt to influence responses. Only 19 percent of Chrysler salespeople asked or begged for perfect scores, compared to the average of 36 percent. Similarly, only 26 percent asked for the chance to address any problems, compared to 36 percent.
Only a single Chrysler buyer reported giving a dealer higher scores as a result of such tactics, and fewer reported feeling pressured (19 vs. 27 percent). Even so, 43 percent gave the dealer
all perfect scores, compared to the average of 36 percent.
Dodge (80 responses)
Dodge buyers were less likely to give the desired perfect scores: 29 percent, vs. 43 percent for Chrysler buyers and 36 percent for all buyers. Otherwise, near the averages.
Ford (108 responses)
Close to the averages.
Honda (193 responses)
Close to the averages.
Hyundai (70 responses)
Hyundai buyers were much more likely to give the dealer below average scores, 16 percent vs. 5 percent.
Consequently, the percentage of perfect scores is lower (29 vs. 36 percent). This despite more frequent begging for perfect scores (20 vs. 8 percent).
Hyundai salespeople were also more likely to mention that their bonus depended on getting perfect scores (23 vs. 9 percent). These tactics did have more impact than usual:
twice as large a percentage of Hyundai buyers reported boosting their scores as a result (24 vs. 12 percent). It seems that the average experience is so poor that even with the bump
Hyundai dealers tend to receive lower scores.
Infiniti (44 responses)
All deviations from the average small enough to possibly be effects of the small sample size.
Mazda (112 responses)
Much less likely to have received all perfect scores (22 vs. 36 percent); instead very likely to have received nearly perfect scores. Otherwise near the averages.
Nissan (70 responses)
Nissan dealers were much less likely to have received all perfect scores (23 vs. 36 percent), and more likely to have received average or lower scores (22 vs. 13 percent). This despite more common requests or even begging
for perfect scores (50 vs. 36 percent). A larger percentage of Nissan buyers reported feeling pressured (36 vs. 27 percent), but they did not report boosting their scores as a result any more often than they average car buyer.
Pontiac (46 responses)
Pontiac buyers were more likely to report giving dealers perfect scores (47 vs. 36 percent). Pontiac salespeople don't appear to have done much to
influence these scores. Though more likely to simply mention that the buyer would be receiving a survey, as opposed to not mentioning it at all, just an average number did more than merely mention it. Similarly, 42 percent of Pontiac buyers were thanked for completing the survey, compared to an average of 24 percent. Other possible dealer reactions were rare or not reported at all.
Saab (43 percent)
Slightly above average scores. Little attempt to influence scores beyond mentioning the survey. Saab salespeople were less likely to ask or beg for perfect scores (23 vs. 36 percent) or to ask to address any problems (26 vs. 36 percent). No Saab buyers reported feeling a lot of pressure to give high scores.
Saturn (72 responses)
Saturn buyers were somewhat more likely to report giving dealers perfect scores (44 vs. 36 percent). Given the sample sizes, Saturn dealers were scored about the same as Chevrolet and
Pontiac dealers. Saturn salespeople differed from the average mostly in being more likely to merely mention the upcoming survey
(51 percent vs. the average of 30 percent), with just 11 percent failing to mention the survey at all (vs. an average of 25 percent).
Other tactics perhaps a bit less common than with the average dealer. Saturn buyers were more likely to be thanked
for completing the survey (35 vs. 24 percent).
Subaru (55 responses)
Subaru dealers were less likely to have received perfect scores (29 vs. 36 percent). Otherwise deviations are small enough to possibly be the effects of the small sample size.
Toyota (198 responses)
Toyota dealers were much less likely to have received perfect scores (23 vs. 36 percent). Otherwise near the averages.
Volkswagen (88 responses)
VW salespeople were much more likely to not even mention the survey (40 vs. 25 percent). Otherwise close to the averages.
Anecdotal evidence suggested that automobile dealers were pressuring and even bribing car and truck buyers in an attempt to earn large bonuses by corrupting the results of
manufacturers' dealer sales experience surveys.
TrueDelta's survey about such surveys found that mild forms of persuasion were widespread. About three-quarters of dealers mentioned the survey to car buyers, and nearly half tried to influence survey responses. However, more aggressive attempts to influence survey responses were relatively rare, and more rare than the anecdotal evidence suggested.
Even mild attempts at persuasion had effects. Over a quarter of all car buyers reported feeling pressured to give perfect scores, and about one in eight admitted to giving the dealer higher scores as a result. Given that dealers were given above average scores by nearly nine out of ten buyers, it's likely that this result under-represents the actual impact of dealers' attempts to influence responses.
Overall, many Japanese and Korean brands receive lower dealer sales experience scores than others. In the cases of Hyundai and Nissan, scores would have been even lower if salespeople were not exerting a higher than average amount of pressure on buyers. BMW salespeople were also more likely to pressure buyers for higher scores, but in their case they are highly likely to actually receive perfect scores, possibly because the sales experience actually tends to be a good one.
TrueDelta.com is the Internet's most innovative auto-owner survey website. With nearly 18,000 vehicles currently enrolled in its Vehicle
Reliability Research Panel, TrueDelta.com is fast becoming the Internet's premier pricing and reliability resource. Consumers can perform vehicles
price comparisons, view Vehicle Reliability Survey results, or join TrueDelta's consumer panel at
Appendix: Respondents' Comments
The survey included a space for respondents to leave comments. Here are some that were submitted. (In some cases these have been edited for spelling or length.) These were the comments most relevant to the survey (most comments discussed the purchase itself, not the survey), but are not
necessarily representative of the typical buyer's experience.
Some people noted that everything truly was perfect:
I would have given them the same score without them asking. They were outstanding to deal with.
They got perfect scores because the saleswoman was that good.
They mentioned the survey and how important it was to them to receive good marks on the survey, and how if anything wasn't absolutely perfect they wanted to know about it so they could resolve it. However everything was absolutely perfect...
I gave them perfect scores because they deserved them...it truly was the ideal vehicle purchase.
Others noted a lack of pressure:
I gave them slightly higher scores precisely because they did not pressure me about the survey.
The dealership wanted me to know that the survey was very important to them. I had no problem with that.
Dealer provided a good deal, good service, and so when the survey came in, I filled it out as I saw fit. They merely requested that I do indeed fill it in and return it to Saab.
They said "At Toyota anything less than a perfect score is failure. So if there is any reason you would not give us a perfect score, please tell us now so we can fix the problem immediately." I liked that attitude and was very happy with the purchase experience.
Our salesman handled the anticipated survey the same way our Project Mgr.s at AT&T/Lucent Technologies did - very professionally.
They deserved the scores. They did an excellent job. I won't say they begged, but they certainly mentioned the survey a few times, including an email from the owner complaining about the survey.
The most common comment was that the dealer said they needed perfect scores, and often asked for the opporunity to fix any problems, so there would be no need to report them:
The ONLY contact from the dealership after purchase was an email saying please give them a perfect score.
The salesman explained how the survey worked (that anything less than "perfect" means he failed). Since I enjoyed working with him I gave him "perfects."
They stated that they would have problems from higher up if they received anything less than "Completely Satisfied" responses. It was a great purchase (knowing and accepting how the car buying process works) experience anyway.
The salesman stated that if I were to give him any low ratings on the survey, to let him know before I returned it.
Dealer didn't pressure me for perfect scores; they said they would criticized harshly if the got anything less than perfect and if I was happy with them to please consider giving perfect scores. I had a good buying experience so I obliged.
Their standard line was "If you can't give us a perfect 10, give us a call and a chance to fix it."
They all say the same thing "Any score that is lower than the maximum score is regarded by [Car company X] as a failing score. Please contact us if you are not able to give us a perfect score."
Even at Saturn they made sure you understood a less than perfect score would be a terrible thing for the salesperson involved, and they say this multiple times. They shouldn't even mention the survey if they really want real responses.
Would have listed several negatives if I had not agreed to post all positive results.
Some dealers provided a template for the buyer to copy or offered to "help" with the survey:
They offered to help me fill out the survey, but I said I could do it on my own...
They showed me a survey with all perfect scores & told me that that's what they expected. They mentioned the survey several times - that it would come in within 4 weeks, that it would come in around the time my tag information came, etc.
A number of commenters were told that pay or even employment depends on the scores:
My salesman told me his employment status relied on getting perfect scores. I did not believe him and marked him lower for lying to me.
The saleswoman told me I would be surveyed and asked that I give her good scores, that it would be beneficial for her. She did not ask me to give perfect scores.....
The salesperson told me their pay/bonus was based in part on the survey, and they encouraged me to let them know if there were any problems. But I never felt that they "pressured" me, per se. They never asked me to be dishonest or to give them perfect scores.
Have had salesmen say they wouldn't get a bonus if I did not give perfect scores EVERYTIME I have purchased a new car over the past 4 years (four cars for immediate family).
The good scores were given out of pity more than anything else.
Some car buyers were offered a gift:
Was told that if I brought back the survey to the dealership and "we completed it together" I'd get a free tank of gas. I didn't take them up on their offer.
My sales rep offered me a free tank of gas if I called him when the survey arrived so he could go over it with me. When my survey arrived in the mail, I called the dealership but my sales rep was no longer with them.
Some comments described especially negative experiences:
You should see the appalling experiences that some buyers have had with this absurd, specious system. For example, on E90Post and Bimmerfest some buyers reported having sales associates personally threaten them for having given scores that were less than perfect.
I would never go back to the dealer because of the pressure they exerted on me to give them a high score which was not an honest response.
Salesman asked me to exclude my experiences with the finance department from the sale. My experiences with the finance department almost nixed the deal and I will probably not buy a car from this dealership again.
I complained about the car having coffee stains on the seat and console when delivered and a minor defect with the interior trim. They did rectify the situation quickly but were miffed that I reported it in the survey as requested by the manufacturer.
The salesman made a huge deal about it and why he had to have perfect scores. I was unhappy to have to put up with this extra stuff during the sale - it takes long enough as it is!!
The sales survey is one of the most nauseating parts about buying a new car. The fact that the salesperson (and sale manager/F&I guy, by the way) have to ask you to answer perfect for all scores tells you there's probably something wrong with the picture.
I was treated VERY badly (yelled at, cussed at) after they saw my survey responses. I wanted to complain to the dealer G.M. or Nissan H.Q. but I was too afraid of more retaliation from the salesman. Still makes me angry to this day, the way I was treated...
Some commenters reported giving perfect scores because they felt that the survey mattered far more to the dealership than it did to the car buyer:
I used to work at a car dealership and I know how much those stupid scores matter. I did have some issues with my new car but they were appropriately resolved. I gave a perfect score because it was worth little to me but I know it is worth a lot to them.
I gave them perfect scores as the dealer really pressed it, and I don't really care, as if I give them bad scores or good, they still won't change their sales tactics.
Some commenters reported not returning the survey in response to the pressure involved, and in some cases possible dealer retaliation:
I never respond to these types of surveys because they always tell me that anything other that a perfect score will hurt the dealership and salesperson. I did have a pleasant experience at the dealership though.
I received an inquiry about my buying experience from the dealer, but I didn't respond. Didn't want to make anyone at the dealer mad (my comments would have been negative) in case I needed them for something later.
My experience at the dealer was mostly positive. However, rather than "ding" them by complaining about the length of time it took to finalize the deal (we were there over four hours), I simply opted not to fill out the survey.
It's an entirely flawed system. My salesman did not deserve perfect scores, but any value less than perfect would grant me a call from them (since it is not an anonymous survey). I opted instead to not complete the survey.
I purposefully did not fill out the survey because they were high pressure sales and requesting I fill out the survey to benefit them, they kept asking me that I was satisfied and that I wasn't being pressured even though I was...
Some commenters reported that they never received a survey, in some cases because the dealer provided the manufacturer with incorrect contact information:
They said I'd get a survey - never got one. I really would have liked one to recount my experience with the sales manager trying to reverse the transaction on me so he could sell my car to the next 2-3 people he had lined up.
The dealership (C____ BMW in [city]) knew I was unhappy so they deliberately gave a false phone number for me to the manufacturer, so that the manufacturer (BMW) would be unable to reach me. C____ BMW people are the sleaziest I have ever met...
Although the salesman gave me a sample with perfect scores marked, I didn't receive the survey. I emailed Subaru of America (SOA) repeatedly and it was finally confirmed that they had mailed my survey to an incorrect address.
Some commenters felt that the process was too corrupt or incomplete for the surveys to be of value:
Salesman asked for perfect scores. Said anything less than a five (perfect) was scored as a zero. The survey, in my opinion, is flawed if it is really scored that way - I rarely think anyone does a perfect job.
Surveys miss the point. I visited 3 other Infiniti dealers and encountered incompetence and lack of interest. Manufacturers should ask about other dealers visited, I think they would get an earful!
I mentioned this coercion in the comments section of the survey. Nothing was said about it to me afterwards, neither from Infiniti corporate or the dealership.
They seem to mention the survey and imply that a lot of their livelihood is tied to the survey (almost a guilt trip). I try to be honest, but on the [surveys] tend to be more positive. I really wish that getting good honest feedback was their aim, not gaming their bonus.
Surveys seem pretty worthless since they are skewed upwards due to sales staff pressures.
The surveys seem to be a joke. I hear from other buyers how much pressure is put on them to only check perfect scores. All need to be anonymous to truly be valid.
Finally, many people mentioned that they receive much more pressure for top scores when they take their car in for service, but that was not the focus of TrueDelta's survey. Similarly, many people wish they were asked as much about their satisfaction with the car itself as they were asked about the sales experience.
No one said, "I'm happy that manufacturers conduct these surveys, because they do a lot of good."