Is this ethical?


Many car makes have adopted what I think is a model first perfected by BMW. Not a car model, a business model. That is, build some base engine, next turbo or supercharge it to create the a performance level differentiation (that difference is legit), then use software boost limits to create several other performance level distinctions. Next, charge customers tens of thousands of dollars to up the boost limit in software. BMW was the first I noted doing this fleet-wide in the early 2000’s, but maybe they adopted the business model from another brand. Jaguar is clearly an adopter, of late.

Question: Is this ethical?

In most cases, customers do not understand the only difference between say, 575hp SVR engine hardware, and 495hp V8S engine is, well, nothing. And they certainly do not understand that the same engine in less expensive form is absolutely guaranteed to last longer, since the only thing they are buying is a different software entry.

I think this practice a scam and should be treated as such. Manufacturer’s who adopt this practice should produce their actual cost delta between programming a car one way, and programing it the other. If their cost delta is not in some way commensurate with their price delta, they should be called out for intentionally scamming customers. Especially when the large price premium inevitably results in manufacturer-known premature engine failure compared to the “cheaper” software implementation.

(Brett) #2

My take on it is they’re giving the people what they want. I fail to see how this is unethical. Fact: people will pay more for higher performance. If that’s done through programming or different hardware really doesn’t make a difference. As for those cost differentials, they’ll get as much premium as the market will bear. Everything is expensive on new cars. Pretty much any option costs no where near the premium charged.

(Gunnar Helliesen) #3

I’m an IT guy. Back in the day, my then employer (an oil company) had a large Control Data Corp Cyber mainframe. We’re talking a large room full of hulking metal cabinets. We’re talking the lights dimming when we had to perform a ‘Deadstart 0’ (power on) of this thing.

We had a whopping 32 MB of RAM memory in our Cyber, but it wasn’t enough for our needs. So we contacted CDC for an upgrade to 64 MB. The $$ amount they quoted us was enough for my employer to elect to lease the upgrade instead of buying it. And this was an oil company, remember.

Once the paperwork was signed, our resident CDC engineer (yes, resident. He was an employee of CDC but he had an office in our IT department) told everyone he was going to shut down the system in preparation for the upgrade.

We thought he was nuts. The ink wasn’t even dry on the paperwork, there was no way the extra memory could have arrived yet. He smiled, calmly walked into the server room, shut down the Cyber, and opened one of the access doors at the back of the memory cabinet. He then proceeded to pull a pair of pliers from his pocket.

You guessed it. Our Cyber mainframe had shipped from the factory with the maximum amount of memory already installed. There was a wire (technically, a wire wrap, connecting two pins) in place to disable half the installed memory. All the engineer had to do was cut that wire, and presto, we had 64 MB. My employer paid CDC a small fortune for this little snip. It was so easy that any one of us could have done it, had we known which wire to cut.

Needless to say, my employer was not amused. But legally there was nothing they could do. Was it ethical? That depends.

When you download a free 30 day trial of a software package from the Internet, that is very often the full product software package you’re downloading and installing. There is, bit by bit, not a single difference from the (often expensive) fully licensed version and the free 30 day trial version. All you have to do is purchase a license, then enter a code into the trial software. Once done, the exact same software package that you’ve already downloaded and installed turns fully functional.

There is one important difference, though. When you buy a car (or a mainframe computer) it becomes your property. You never buy software, you license it. You’re buying the right to use it, not the property rights to the software itself (with a few exceptions).

The reason I bring this up, is that the mindset of the car manufacturer is the same. In their minds, they sold you the V8S. You didn’t just purchase the hardware, you purchased the driving experience, a posted set of specs, and the status that accompanies the specific model you purchased. In other words, in addition to the hardware, you purchased a bunch of intangible goods as well.

By limiting horsepower in software, the vendor will argue that they have taken nothing away from you. You purchased, and now own, 1 engine block, 2 heads, 8 pistons, and so on. If you want the status and the driving experience of the higher priced SVR, you need to pay up.

It’s only a question of time before car manufacturers implement the equivalent of HDCP* in cars, where performing an aftermarket software upgrade (or “hacking”, as I’m sure the car manufacturers will call it) will be impossible. Tesla is already quite far down that road with their EVs. Tesla owners are already asking who owns the damn car, Tesla or them?


*) HDCP = High-bandwidth Digital Content Protection. It makes sure all components in a video system are approved to handle copyrighted content. If not: No picture. Car manufacturers can do the same. If you install software that’s not approved by the manufacturer: All other computers/controllers in the car will shut down and immobilize the car. The same could apply to aftermarket (hardware) accessories. If it plugs into the electrical system, the manufacturer can verify that it’s approved and take action if it’s not.

(Robert Wilkinson) #4

Wow, folks now days don’t realize how whopping that is. In the early 80s, our DEC computer maxed out at 128 KB, and that was about 8 1ft x 2ft cards jammed with dynamic RAM ICs for a total cost of $8-9K IIRC. So the cost of ~20 times that much memory would be up around $100K! Our discs were only ~20MByte.

It’s remarkable that they put that extra memory into your Cyber essentially on consignment, assuming you owned rather than leased.

It’s clear that the current business model for cars, like electronics, is heading towards de facto leasing rather than owning. Software is the most obvious–you don’t even think you own it.



That’s a great, similar scenario that we are all familiar with. One thing is different, though…

In your IT example, the “more expensive” solution presents no downside to the buyer, only to the vendor. IOWs, the vendor absorbs the higher cost of the additional unused memory capacity on the low end of the sales curve, presumably because it is cheaper for the vendor to include, rather than omit memory, compared to configuration controlling their product line.

In the case of a car manufacturer, at least one major, undisclosed part of customer value delivery is opposite the price delta: design life.

The “higher performance” car buyer gets more power at the direct expense of known, lesser design life. This is 100% understood to the manufacturer, but not necessarily known, and certainly is not disclosed to the customer.

I don’t see anyway to argue that extracting significantly more torque and power from the same engine hardware does not equally-significantly lower the known engine design life.

I’m sure JLR might argue that the customer has opted to pay more for the additional performance, fully understanding this comes at the direct expense of shorter design life since they do not hide otherwise identical engine specs. They might say the customer even enjoys this trade as it gets them into a new car, sooner. That, IMO, is nothing but rationalizing known unethical behavior at the point of sale.

In no way do manufacturers disclose the shorter lifecycle specifications of hopping-up an otherwise identical engine–unless the customer does it to them, of course. Then alarm bells ring and warranties are cancelled.

(Gunnar Helliesen) #6

Good point, and you’re absolutely correct. But, consider this: I wonder how significant that design life delta really is?

If I’m buying a new car every 3 years (I’m not, but bear with me here), and the second hand value of the higher output engine remains good and higher than the value of the lower output engine, why would I care?

If I purchase the car to keep it for its entire life expectancy (say 20 years, 200k miles, and beyond that as a classic car/Sunday driver), will I have to rebuild the higher output engine sooner than the lower output one, and how much sooner? If the answer is after 18 years vs. after 20 years, then again, I may not care. If the answer is after 3 years and 1 month vs. after 20 years, then I’ll care a lot. And loudly.

How much will my individual driving style (aggressive/relaxed) affect the answer to the above question, vs. “built-in” life expectancy (increased wear from higher pressures and higher temperatures etc.)? If my personal driving style is a much higher predictor than the built-in life expectancy, I may not care enough to change my driving style.

Either way, it would be very interesting to see some numbers on this.

(Erica Moss) #7

Is it within the power of the customer to flash the eeprom? I suspect so because they are required (I think) to provide that information to third parties. My VW was flashed to give a roughly 30HP boost. I understood the implications of doing so after discussions with my mechanic. I could have gone much farther with it but was told it would have required strengthening of certain components.

So you might have a legitimate beef here if manufacturers know that a higher tuned engine is more liable to fail and there is no disclosure at all to customers. I don’t think they need to disclose the cost of flashing though any more than they need to disclose the cost of any other manufacturing step. It isn’t just the programming of course but also the R&D and testing involved.


Definitely. And in this sense the (IMO unethical) practice offers adventurous customers a unique opportunity as well as a pricier option. I bought my BMW 135 specifically because I knew the engine was capable of much more than the 306 hp, as sold. Around $1000 later–$500 for a piggyback ECU and a few bolt ons–I have a 420hp car. That’s 100hp more than a the M version and obtained at ~$20K discount. I bought my 495hp V8S for the same reason. But… both BMW and Jag will cancel their warranty if they find out you have removed boost limits, even if it is essentially identical to what they sell for way more $.

Here’s food for thought straight from the Kelly Blue Book Website…

2017 Jaguar F-Type V8 SVR (575hp) new: MSRP $129,795 (base)
Current used value with 20K miles: $78,763 (base)

2017 Jaguar F-Type V8 R (550 hp) new: MSRP $$109,245 (base)
Current used value with 20K miles: $68,747 (base)

So the market dictates exactly twice the depreciation rate for +25 hp, even after some pricey carbon fiber upgrades and still in warranty. Will the SVR sell for the same or less than the R as warranty fades to 0? I think so.

(Brett) #9

Am I understanding you right? Did you say that the SVR’s market price will dip below the R’s market price and sell for a discount to the R? The prices will move towards convergence over time, but its never going to invert.


Often, more expensive cars sell at a big discount to less expensive cars from the same line, because the market understands the anticipated repair bills after the warranty expires. This is especially true with high performance cars in the $100-$500K MSRP range with 8-12 cylinders or difficult mid-rear engine access, as commonly needed engine repairs easily exceed the residual value of the car.

Here’s a $200,000 Mercedes 604hp/740tq 200mph V12 Bi-Turbo selling for less than the same year Corolla…

So if the 575hp SVR has more problems than the identical engine producing 495hp in a V8S or 550hp V8R, it could definitely sell at a discount. Consider that Jaguar will cancel your warranty if you add the same 25hp to a V8R. There’s probably a reason for that.

That’s my whole point here, they never disclose the downside of the same hardware running harder. Most SVR buyers probably don’t realize thast $20K performance premium buys them an absolutely identical engine.

(Nick Saltarelli) #11

In the early mid 80s, when I was a Registered Professional Forester, I was asked to provide an estimate of one hundred year wood supply for a 1.2 million hectare tract of forest. The only way to pull it off was computer modelling so I requested and was granted a user ID and password. The company I worked for had at the time employed an IBM 3081 mainframe running System/370 architecture. This was located in Toronto head office and each of the thirty or so divisions across Canada and the US could access it through a remote terminal, or virtual machine. The JCL was CMS EXEC and the most commonly used language modules at the time were 4th generation like dBase and C+ but earlier generation languages like Cobol and Fortran were still supported. I built a model using Fortran IV using the multiple-million lines of forest inventory data as its base. The model was several thousand lines within which were imbedded hundreds of logical tests and GOTOs but the thing that prompted the sytems VP to hail me on the phone to chew me out for crashing the mainframe was the memory assignment I’d allowed in the first few lines - including seven dimensional variables requiring 10 Mb of reserve.

“TEN MEGABYTES?!”, he fairly yelled over the phone, “ARE YOU NUTS?!”

Given the complexity of the assignment and enormity of the data base, I was finally allowed to have five, though not without some grumblings from IT.

(Robert Wilkinson) #12

Slightly drifting, but I’ve been wondering if anyone else shares my perhaps nonsensical feelings about memory and, a closely related concept, bandwidth.

In my youth, early and mid-adulthood, memory and bandwidth were expensive commodities. All of my experience was with memories of around 4K (the LINC, first lab computer, later called PDP8) to 1Mbyte, and tape (later disc) storage of perhaps 2-20Mbyte (per disc or tape). Later, my work focused on imaging (a lot of 3D and 4D, meaning 3D movies) and I was up in the gigabytes. Throughout this time, memory was expensive (mine paid for with US govt grants), as was the means to transmit data (bandwidth).

My lab routinely purged unneeded images and movies to save space and money. Same for other types of data, and for source code/object code that would no longer be used. Nothing extra loaded on computers, of course.

Now, jump to the present. My desktop has a lot of junk on it, including Windows applications, not to mention my own unused stuff. The OS wastes tons of memory, a trend that continues with each upgrade. Every day, I get several iPhone images or movies from friends and families–at least 1 of our dog Merlin per day included. I can’t even figure out how to get rid of this stuff from my phone without some effort. And I’m not sure the effort is worth it…it seems like the memory is limitless. Much of this has to do with the relatively recent invention of flash memory, of course.

Another example is Jag-Lovers. Whilst on the old platform, I eagerly anticipated the day when I could post photos of my completed car. It never came…car still not presentable in terms of the photos on the old site to wast the storage or bandwidth. Now, with the new site, we are all posting movies, images or links just for the fun of it. I feel guilt about this too. What if memory and bandwidth become expensive again?

Anyone else feel guilt for using excessive memory or using excessive bandwidth? I’m sure younger folks have no idea what I’m talking about.

(Paul Wigton) #13

Not in the least: that said, if it gets expensive enough for JL, Ill happily pay a surcharge to help buy what’s needed.

(Brett) #14

Sorry Sandy, but I just don’t think this bears out in real life. If it did, everyone would be choosing a plain jane 5 Series over an M5 or a six cylinder Mustang over a Cobra. Prices certainly don’t reflect that happening. The only time this applies is for stuff like an Audi A8 W-12 or something that is incredibly difficult to work on and practically experimental. Besides, your point was not referencing emerging technology but instead the same engine with different tunes.


Sure, the market dictates the answer anyway, not our opinions.

Also relevant to this topic, another practice is emerging all over the auto industry that is pretty similar, only worse IMO, since there is not even a performance difference. That is, an advertised hp increase from doing absolutely nothing but re-drawing the red line, higher.

HP = TQ * RPM / 5252

Ethical? Similar? The same thing?

Yes, you can buy two identical cars today, but the newer model has more advertised HP due to nothing but a red line increase. I think this is unethical too. Technically, the cars are identical and even have identical performance (though there might be a software rev limiter involved). Again, the upside is advertised, but the inevitable downside is not disclosed.

Is chipping away at durability the new performance profit center?

This is exactly why physics doesn’t allow using RPM as a proxy for distance traveled in the fundamental [Power = Work/Time] equation, where Work means actually moving the mass of the car some actual distance. But all manufacturers simply ignore the physics required to accurately rate horsepower, because it is so much easier to invoke mathe-magic than to do meaningful work.