Why Am I Making Such a Fuss About Estimates?

Part 1 What's The Problem?

I suppose most people don't understand why I make such a fuss about repair estimates. How difficult can it possibly be to tell a customer how much a particular repair on their car is going to cost?

I guess, from the perspective of an average consumer living in a perfect world where unicorns gallop across rainbows and shit daisies, there is nothing difficult about it. Just say how much. What's the problem?

Well, in that perfect world, all cars are, all cars are basically the same underneath. They differ only in size, body shape, and interior trim. Just like people: our faces and body types are different, but our anatomy is basically the same. Surely cars work the same way. All cars are also in essentially the same condition. With age, they merely acquire the automotive equivalent of wrinkles: get some scratches, dents, and perhaps start farting at night... I meant to say, leaking oil here and there. All those different parts are priced the same for every make and model and remain constant forever.

And, of course, every repair shop has a giant price list containing the price of every possible repair. So when you call a shop and tell them you need to replace the rear shocks, they just look up the line that says “Rear shocks replacement — $350” and give you an answer. After all, when you go to a store, everything has a price tag on it. Nobody tells you, “We need to look at you first and do some calculations before we can tell you how much this jar of mayonnaise costs.”

Therefore, if an auto shop wants to “inspect” your car before giving you a price, they must really be trying to figure out how much money you have in your wallet and how much harder they can rip you off. They're all crooks anyway. So you'll just go to an honest shop — the one that can immediately tell you, without hesitation, exactly how much it costs to replace your starter and fix fault code P0174. They didn't even ask what car I have; they must be the best!

Sounds funny, right?

Unfortunately, I suspect quite a few people won't even understand what exactly is funny about it...

Strangely Enough, the Industry Isn't Much Better

As absurd as this sounds, the automotive industry's management systems often seem to live in a fantasy world, only of a slightly different kind.

At least that's the impression I get when I look at something like a MOTOR parts and labor database, where labor times and parts prices are conveniently presented together. The apparent idea is beautifully simple: take the listed price of the component, take the listed labor time and multiply it by the shop's labor rate, add the two together, and voilà — the repair estimate is ready!

Anyone even remotely aware of the present economy should already see one problem. The listed parts prices are largely mythical. Real-world parts prices jump around constantly and, somehow, seem remarkably reluctant to jump downward.

More sophisticated shop-management systems solve this by integrating directly with participating parts suppliers, allowing current prices to be pulled straight into the estimate. And yes, that can work. However, it also ties the shop estimating system to a limited selection of participating suppliers, their inventory, their brands, and their prices. That limits what I can offer my customers and can restrict either my profit potential or my ability to find the customer a better deal.

But those aren't even the biggest problems.

Problem #1: Where Is the Rest of the Repair?

No estimating system is going to hand me a complete BOM — Bill of Materials — containing everything actually required to perform a particular repair.

Here's a real example. I look up a head-gasket replacement for a 2014 Ford Escape in MOTOR. It gives me the price of the head gasket and a labor time.

Wonderful.

Except this is supposedly a professional automotive estimating system, and it hasn't even bothered to mention the cylinder-head bolts, never mind the dozens of other parts, seals, gaskets, fluids, fasteners, and materials that may be required to complete the repair. We can start with something as exotic and difficult to predict as engine oil and coolant. The last sentence was intended to be sarcastic, by the way.

Apparently those are just little insignificant details for an estimating system.

The software pretends it knows what a head gasket costs. It does not necessarily tell me what I need to perform a head-gasket replacement. Those are two very different things.

Problem #2: The Sacred Book Time

Then we have labor time. This is presented as though it were some predetermined physical constant, carved into stone tablets somewhere. Yes, this repair will take exactly this long. Trust me, bro!

Arizona or Ontario? Doesn't matter. Fifty thousand kilometres on the odometer or 250,000? Clean and garage-kept or fifteen Canadian winters underneath it? Previously untouched or already “repaired” three times by somebody armed with an impact gun and a tube of RTV? Apparently irrelevant.

Head-gasket replacement takes this many hours. The Book has spoken.

Problem #3: Warranty Magic

And then our MOTOR example gets even better. For this AWD Escape, MOTOR lists 17.3 hours. But right beside it, on the very same screen, we have a Factory Time of 12.4 hours.

Same vehicle, same head gasket, same basic operation — and almost five hours disappear.

Apparently, when somebody else determines the factory or warranty allowance, the laws of space and time change. Perhaps the technician gets issued a time machine together with the warranty repair order. Hopefully his life expectancy is extended by five hours as well, just to keep things fair. Take enough warranty jobs, and you'll live forever, man!

Of course, there is a serious point underneath the joke. These numbers may have been established for different purposes and according to different methodologies. A factory or warranty allowance is not necessarily the same thing as an independent labor-guide time. But that's precisely the point: there isn't even one sacred number.

Yet the industry routinely treats “book time” as though it were an objective measurement of how long a particular repair will take. It isn't. The repair hasn't happened yet. Nobody knows at this point how long it will take.

So How Does This System Work at All?

What's remarkable is that, in much of the industry, “management” really does produce estimates more or less this way.

And I put management in quotation marks deliberately. Clicking labor operations in software, transferring parts from an electronic catalog, and communicating the resulting number to a customer isn't necessarily management. At that point, a person in that "management" position functions as a biological interface between the software on the computer and the customer on the phone.

Someone still has to figure out what the repair actually requires, and very often that someone is the technician. The technician is expected to determine and list what parts and materials are actually needed — frequently as part of a “free inspection” — and then gets another interesting responsibility: compensating for the difference between sacred book time and physical reality with his own labor.

If the estimate was wrong and the job takes longer, guess who absorbs it? If something necessary was overlooked, guess who deals with it? And if the technician misses some little gasket that had to be replaced, the vehicle later comes back leaking, and the customer notices and complains, the shop may lose the cost of another gasket while the technician loses several hours redoing the job for free.

Meanwhile, the person who created the estimate can remain remarkably well insulated from the consequences of an inaccurate estimate. What a beautiful system!

But Don't Worry — There Are Gravy Jobs!

The theory, of course, is that labor times contain enough allowance that a “good technician” will eventually learn to beat them. Beat the book. This is a particular source of pride in the industry: I beat the system, I made eight hours in five, I made money, I'm damn good!

And sometimes that's perfectly legitimate. Experience matters. Good organization matters. Having the correct tools matters. Discovering a more efficient method matters.

But this system also creates the famous gravy job: a repair or maintenance work whose labor allowance, for whatever reason, is generous enough that it can routinely be completed faster by almost any mechanic. Sometimes you beat the time because you're genuinely skilled and focused. Sometimes because you've discovered a better method. And sometimes because you learn which corners can be cut without anyone noticing immediately.

When compensation rewards speed directly, speed has a remarkable tendency to become more important than everything else.

I've seen enough of this tendency that I wouldn't willingly send my own vehicle to a shop operating that way, and I wouldn't hire a technician into my shop who is already conditioned by that mentality. I spent only about three years working under that system myself, and it took me probably twice as long to get rid of some of the bad habits I picked up from it.

The Technician as a Shock Absorber

And this is ultimately why the conventional system can produce estimates so quickly. It doesn't actually eliminate the uncertainty. It transfers the uncertainty downstream.

And downstream sits the technician.

The system works as long as technicians are willing to function as human shock absorbers, absorbing its estimating errors in the hope that enough profitable gravy jobs will come along to compensate for the losers.

That arrangement creates its own hidden economy inside many shops: arguments over who gets the gravy jobs, favoritism, resentment, deal-making, and what is commonly called dealership politics. Who gets the gravy? Who gets the nightmare warranty diagnostics? Who gets to be the “productive” superstar, and who gets used as disposable capacity?

Even management skill gets judged by a shop owner’s or manager’s ability to juggle gravy jobs and “bad” jobs between technicians so that everyone remains reasonably happy and everyone gets what feels like a fair share of the overpriced work.

All instead of fixing the obvious problem: labor times that were estimated incorrectly in the first place.

Well, I Don't Have a Human Shock Absorber Here

This system doesn't work in my shop for one very simple reason: the “management” and the technician are the same person. Me.

There is nobody downstream to absorb my estimating mistakes. If I underestimate the work, I deal with it. If I overlook a necessary part, I deal with it. If reality bears little resemblance to the labor guide, I'm the one standing underneath the vehicle discovering that fact.

And if I eventually hire someone — or, more likely, train someone — I have no interest in importing this entire system of gravy jobs, bad jobs, favoritism, and shop politics along with them. I don't want repairs divided into good jobs and bad jobs according to how easily somebody can beat the clock.

A repair should be performed because it is necessary and economically sensible for the customer, not because somebody discovered that it's an especially convenient place to print money.

And that brings us back to the estimate itself.

An Estimate Is an Engineered Product

An estimate is not a price lookup. It is an engineered prediction of a repair that has not happened yet.

The arithmetic behind the estimate is easy: parts plus materials plus expected labor. Anyone with a calculator can do that. The difficult part is determining which parts, which materials, which operations, and how much labor actually belong in that calculation for this particular vehicle, in its present condition.

That is the work involved in producing an estimate worth anything.