TrueDelta

Last May TrueDelta conducted a special survey, asking people what, if anything, car dealers did to influence their responses on manufacturers’ sales experience surveys. Were tactics like asking for top scores, giving gifts in exchange for filling out the survey at the dealer, and so forth as widespread as anecdotal evidence suggested?

Well, the results can now be viewed here. The most aggressive tactics turned out to be rare, but others were widespread. Many car buyers resent the pressure put on them to provide less than truthful responses. Others figure that the survey doesn’t matter to them personally, but it matters a lot to the dealer, so why not give them the scores they need to get paid?

Auto industry trade magazine Automotive News covered the results in today’s issue (sorry, subscription required to view the article). They asked a few manufacturers for a response. Hyundai provided a typical response: “We try not to influence the outcome of the survey. This has been clearly communicated to all our dealers on a regular basis.” Yes, and I clearly communicate to my kids on a regular basis that they should keep their rooms clean.

One manufacturer comment deserves a response. BMW spokesman Jan Ehlen responded that the number of BMW buyers included in the poll was too small to be meaningful. While it is true that only 46 BMW owners responded, the differences noted were so large–about 20 percentage points–that these particular results are conclusive even with the small sample size. When 46 people respond, and not a single one gave their BMW dealer even average marks, much less below average marks, it’s saying something.

Dealers can’t really be faulted for these practices. In many cases, they and they salespeople only earn large bonuses if they receive perfect scores on these surveys. This is an unrealistic expectation. If people need to cheat a system to earn enough to eat, they’re going to cheat it.

In recent years, I’ve noted an increasing number of “Moon and Tunes” and “Sun and Sound” option packages, especially from Detroit. These usually include a sunroof and an optional audio system, perhaps with a CD changer or satellite radio also tossed into the mix, at a discounted package price.

I say “usually” because in the Chevrolet TrailBlazer you can substitute a ceiling-mounted entertainment system for the sunroof–they cannot be ordered together–and thus get a “Sun, Sound, and Entertainment” package sans sun. 

This oddity aside, is there something about a sunroof that makes it a natural fit for superior audio? When these are available separately, do people tend to order both anyway?

Or do “Moon and Tunes” and “Sun and Sound” simply have a nice ring to them?

My guess: it’s just a matter of time before automakers rediscover that “show” rhymes with “go.”

People often ask me why results for each model are not broken out by powertrain. For example, the VW Jetta, Rabbit, and GTI can be purchased with either a five-cylinder gas engine, a four-cylinder turbocharged gas engine, a four-cylinder turbodiesel, or a V6 gas engine. Transmissions include a manual, an automatic, and a DSG. Shouldn’t repair rates be expected to vary based on which of these a car has?

It helps to look at the actual data. Repairs reported for the 2005.5-2006 Jetta (57 cars):

• satellite radio quit (two reports, now a TSB)
• power windows failed
• missing bolt on door skin
• broken rear seat vent deflector
• both headlights burned out
• electrical outlets in center console
• loose door panel causing rattle
• windshield washer hose broke (two cases)
• lumbar support broke
• more rattles
• steering box cracking sound

Repairs for the 2006.5-2007 Jetta, Rabbit, and GTI (31 cars):

• airbag light
• window seal
• seatbelt retractor
• interior hatch trim piece misaligned
• rattle
• compass stopped working
• visor light
• sunglass holder woudln’t stay closed
• bad coilpack (the only powertrain-related issue in the bunch)
• A/C compressor 

Out of both lists, we have a single powertrain repair. All of the other repairs involved components that are common across all variants of these models.

In this case as in most cases these days, powertrain problems have been rare. And the more rare something is, the larger the sample size you need to accurately measure it. To measure problems as rare as these seem to be, you’d need a very large sample. And how meaningful is it, anyway, to report failure rates well under ten percent? If one engine requires repairs two percent of the time, and another requires repairs three percent of the time, it is a difference of 50 percent. But we’re still talking about one extra powertrain repair per 100 cars. Is this really important to know when buying a single car?

If a particular powertrain ever is actually trouble-prone, it’ll show up even with TrueDelta’s current sample sizes, and I’ll note it on the results page. There are a couple cases where I’m considering this currently.

The TrueDelta Vehicle Reliability Survey results suggest that the repair rate of the new for 2007 Infiniti G35 is high, about 79 repair trips per 100 vehicles per year. However, while this is the initial repair rate when annualized, in future updates the reported repair rate will likely be lower.

The reason: nearly every repair reported was for one of two items: a non-functioning passenger seat heater or detached sunroof trim. Infiniiti has a TSB for both. This means that they’ve recognized that these are common problems, and have a fix for them. This fix is no doubt included on more recently produced G35s.

This is standard practice for automakers, and explains why repair rates are often higher for the first few months after a new design is introduced.

Going forward, the Infiniti G35s that have been fixed shouldn’t have these problems again, while newer cars will probably never have these problems. So, unless new problems pop up, the reported repair rate in future updates will be lower, even much lower. (Though there might be a brief blip this fall when people first try to use their seat heaters.)

Stay tuned: the next update will be in November.

No part of TrueDelta’s Vehicle Reliability Survey (with the exception of email delivery) causes more trouble than the process of reporting a repair trip where parts were ordered and then another when the parts are installed. Often people only report one and not the other. How should this be handled when analyzing the data?

To date, analyses have only included repair trips that included a successful repair. So if the only outcome was that parts were ordered, the trip was not included. This caused two problems:

1. Some people forget to report the second trip, probably because they remember already reporting the problem.
2. Sometimes it takes a long time–nearly a year in some cases–for the parts to come in and be installed. Meanwhile, the owner is living with the problem.

In both cases past analyses risk underestimating repair rates. So, starting with this month’s results, repair trips where parts were ordered will be included in the analysis until the parts are installed, at which point only the latter trip will be included. This doesn’t affect many models, and most of those it does affect it affects only a little, but it nevertheless makes the results more accurate.

The revised analysis also will warn of high repair rates more quickly with affected new models. When a model is new, parts for common problems are often on backorder. And even when parts are available, the repair trips for newly introduced models tend to only involve ordering parts, with the parts installed a month or two later. The vehicles simply haven’t been around long enough to go through the typical cycle of going to the shop, ordering parts, then returning to the shop to have these parts installed. With the old method, the results wouldn’t reflect these problems until the parts became available and were installed. With the new method, problems will have an impact on the reported repair rates with less of a delay. 

There is one other benefit. Most of the responses I’m still waiting for from participants involve open issues. Hopefully these members will still respond. But in nearly all cases they’ll report that the repair was completed in a previously unreported repair trip. The revised analysis anticipates this response, allowing preliminary results to be posted now.

From time to time I receive an email asking which cars are most comfortable for someone who is tall, long-legged, large, or all of the above. Problem is, I’m none of the above–at 5-9 and 175 pounds I’m probably very close to the size person cars are designed around. So this isn’t something I can evaluate well during my test drives. Even so, I’ve long noticed that the official interior dimensions provided by manufacturers are at best a loose indicator of how roomy a car actually is. Why? I don’t know, but that’s the way it is. So the specs are nearly useless for answering such questions.

In the end, the best information will come from people who are tall, long-legged, or large and who have paid attention to comfort while shopping for their own cars. If you fit any of these descriptions, please either leave a comment or contact me (see link at the bottom of the page). Whenever possible, be specific about how a particular car excels (headroom, legroom, cabin width, ease of getting in and out, and so forth). I’ll update this blog entry as information comes in.

I strongly encourage people of all shapes and sizes to evaluate roominess and comfort as much as possible before buying a car. Too many people discover that a car is physically uncomfortable for them after buying it–and this is something that the warranty and lemon laws do not cover. In other words, once such a problem occurs, you’re stuck with it as long as you own the car.

I’ve been reviewing the 800 or so repairs reported over the past quarter, and once again one problem stands out as the most common. Beginning next month, all vehicles will have to be equipped with a standard tire pressure monitoring system (TPMS). Since last year 70 percent of all vehicles have been required to have such a system as standard equipment. Problem is, these system have a high failure rate in many models, including the Nissan Versa. So if you have a problem with the TPMS in your car, you’re far from alone.

If memory serves, the TPMS mandate was a response to a number of fatal 1990s-era Ford Explorer rollovers, many of which were partly caused by underinflated tires. Many people don’t check their tires often enough, especially in this age of self-serve gasoline. So now everyone gets a yellow warning light when a tire needs more air.

There are two major types of systems. The more precise solution is a pressure gauge in each wheel that wirelessly communicates with an on-board module. The most sophisticated systems go beyond a warning light to provide the actual pressure reading for each wheel. The cheap solution uses the ABS sensors to compare wheels speeds, in combination with just a pair of direct pressure sensors (in case all four tires are equally underinflated). An underinflated tire has a smaller diameter, and so it rotates faster. Some manufacturers argued for an ultra-cheap system that used only the ABS sensors, and such systems were used in some cars before the mandate went into effect, but the government rejected this as a sufficient solution.

Problem is, in many cars the yellow light comes on–and stays on–even when the tires are properly inflated. When the TPMS isn’t working correctly, sometimes only a system re-set is needed. But other times a sensor or a module must be reprogrammed or even replaced. Currently the former two solutions aren’t counted in TrueDelta’s reported repair trip rates, but the last is.

When all else fails to fix the problem, some people resort to a square of black tape. And, no, that doesn’t count as a successful repair.

In late June I asked, “Satisfaction studies, is there a point?” Well, today the results of another such study, by Claes Fornell of the University of Michigan, were released. And The Detroit News subtitled an article based on these results “Customer contentment with U.S. automakers improves sharply as Asian brands drop.” Really? Let’s take a closer look.

Results were only released at the brand level, not at the model level. And as with many results at the brand level, these don’t vary much. All but one brand falls within five points of the average, and 13 of 21 brands fall within a mere two points. This is two points out of 100.

And the sharp improvement reported by The Detroit News? A couple of domestic brands were up by three points compared to last year, and a couple others were up by two points. The others were up by a point or less. Jeep even feel by two points. As did Honda and Nissan. While Toyota feel by three points.

But the real news is that the differences, whether from brand to brand or from year to year, are almost all small. No brand changed by more than three points compared to last year. And even with sample sizes of over 100 for each brand, a point or so is bound to be random variation or simple rounding. So I don’t see any sharp changes.

To its credit, The Detroit News did note that “the gap between the highest- and lowest-performing brands has been cut from 18 points to 12 points since 1994.” Problem is, people want to read about dramatic changes, not about a lack of differences. So journalists are under pressure to take two- or three-point differences and write as if these were far larger.

Each quarter, after data collection ends, I send emails to those with a gap in their responses seeking to close said gap. This time around, these emails went out to 50 members who responded for June and in earlier quarters, but who missed either December or March. Just a couple percent of the total, but still well worth doing.

For example, take someone who responded in September, December, and June, but not in March. Without filling the gap, only September through December and the month of June could be included in the analysis–five months. Once the member fills in the gap, usually by communicating that no repairs occurred during it, January through May can also be included, five additional months. So a single simple response doubles the amount of data for this particular vehicle.

Aside from increasing the amount of data, filling the gaps yields more accurate results. Rarely does someone report a repair in response to these gap-filling emails. It seems that when people fail to respond, it is because they had nothing of obvious significance to report. But it is just as important to know if there have been no repairs. If the gaps were left unfilled, the resulting repair rates would be a bit higher than they should be. So we strive to fill as many as possible.

Data collection is now complete for June. I held the data collection open for a few extra days, in an attempt to include a handful of models that were a single response short, but to no avail. These included the 2004 and 2005 Cadillac CTS.

Even so, the August results will include 52 models, up from 33 in May. There will also be another 44 asterisked results with sample sizes between 15 and 24, for a total of 96 results. This is up from 65 in May.

The response rate was two percent lower this time, at 41 percent. The site was down for an entire day recently, which could explain much of the difference. TrueDelta will probably be switching to a new host later this month because of Apollo Hosting’s increasingly poor performance and abysmal support.

I now need to go through the data, fixing incorrect or incomplete responses. Once this is done, the latest results will be posted.