Rooster's Revenge

Just a quick note to say hello to all the Die Guys out there!

All evidence to the contrary, I have started writing again. I teamed up with my pal Lou Kren to collaborate on an article for one of his ventures. I am my toughest critic – but I am very pleased with my work with Lou.

I also have no less than 5 book concepts in the works. At the rate I am going, it will take a couple years to get it all out there – but am close to getting the first one going. It will most likely be an engineering calculations-focused book.

I feel good overall – and I genuinely believe I am doing my best work ever.

That said, did you know that Vince Lombardi prepared his entire career to be a head coach? And no less than 20 years into his journey, he finally got the opportunity with the Green Bay Packers and helped put the NFL on the map. [my prediction is they will be in this year's Superbowl and walk away with the Lombardi trophy]

We are approaching the 25 year mark on my career. Does my dream job of building a championship-caliber die engineering and build team with owning the quoting method and die standards lurk on the horizon?

If I had THAT opportunity, the possibilities are boundless. I foresee a reversal of work going overseas back to my hometown with the right die engineering and build processes with my brand of modern die standards to guide the efforts.

I do not know if 2011 is my year or not, but I know this: I am having fun right now.

The curse of technical people, I am convinced, is they are too technical in situations where detail does not matter. They have an obsession with covering every minute detail when a high level concept is all that is needed.

Case in point: advanced feasibility of metal stampings.

Here is a situation where the product geometry, material type, and material thickness is far from frozen. “Frozen” is an automotive term for final designs that are released for production. Final, in this market, means “design intent will most likely change unless hell freezes over”.

At this point in the product development process, someone somewhere is simply looking for feasibility. Not validation.

The formability and proposed process is all we are looking for here. Is it a crash form or a form or a multi-stage draw?

Will it run in a progressive die or transfer?

Hell, even a rough blank size and budgetary tool cost may be required as well. Rough does not mean down to two place decimals on dimensions. Budgetary does not mean down to plus or minus 2 cents.

And that is my two cents: getting down to the level of detail of sourcing blank suppliers, getting 57 people to sign-off on a process that stands a 90% chance of never turning into a job for a product design that stands a 99.9% chance changing somehow someway to make all the advanced work a complete waste of time.

Focus on what matters. Don’t go overboard.

Interesting conversation the other day with someone about getting the job done fast. Although the topic of discussion was quoting dies in advance of an actual design, this applies to design, formability analysis, and just about every other J-O-B in this business.

It seems like everyone rushes to say “we need faster quote tools” or “we need faster simulation software” or “we need faster die design programs” and the like.

I counter with faster does not mean better. Better is usually accuracy. Believe me, they want accuracy too. But speed is the challenge.

At the heart of the speed challenge is usually not about the job of quoting, forming, or designing. Quite the contrary.

When I was doing die design, I spent more time seeking information to do the job than it took to do the design job itself. The same goes for when I did formability analysis. Hell, I would spend more time trying to find out what some bullshit proprietary commodity code for material really was than it took to set up and run the simulation.

So, let’s take quoting for a detailed breakdown of what I am talking about. To quote a job, you need a process. Once you have a process or die operational lineup, most companies use some form of semi-automated Excel spreadsheet for quoting.

To process and quote a job, for instance, this is typically what happens:

1. Search for CAD file. (5 minutes)

2. Launch one-step to get blank size. (5 minutes)

3. One-step needs IGS and you have a Catia file. Go get coffee. (10 minutes)

4. Launch Catia, open file, and save as IGS. (10 minutes)

5. Run one-step to get blank. (5 minutes)

6. Determine process (5 minutes)

7. Get files needed to create process and quote forms (5 minutes)

8. Export images of part and blank from one-step (5 minutes)

9. Import images into process sheet and crop so they look pretty (5 minutes)

10. Fill out process form by copying info from a BOM (9 minutes)

11. Use Ouija board quoting tool to get price (1 minute)

12. Put quote 10 folders deep on a server to make them hard to find (10 minutes)

13. Take the quotes out of the folder and send to 57 people for review (10 minutes)

14. Make necessary changes and put back in the folder (10 minutes)

15. Take a nap until the next job comes up (usually 5 minutes) 

As you can see, most of what needs to happen to get the job done is NOT time well spent. Instead, it is spent doing things that really do not, or should not, matter.

Bottom line in this example is the quote itself – the price – is only 1% of the effort. Getting the quote depends on the process and the process takes 9% of the effort. The other 90% is spent getting to the point of doing the meaningful work.

Let’s say you cut the time spent quoting and processing by 90%. This comes down to 1 minute. Everything else still takes 90 minutes (in this example anyway). So now instead of having 90% of non-value, we have 99% non-value to complete the task.

Why not focus on cutting down on the non-value like searching for CAD files and material specs and such? Keep the coffee and lose the folders I say.

If you want to go faster, focus on cutting out the steps that consume non-value added time, like finding and converting CAD files.

My view of dies is their architecture should be driven by both the production volume and material strength of the stampings. Unfortunately, this is fuzzy in practice. Many die standards and quoting systems do not take both into account.

The illustration below captures how I believe dies should be built based on the combination of production life and material strength:

There are three main categories of architectures: Low, Standard, and High. It works like this: Depending on the combination of production volume and material tensile strength, the low and high architectures are variations of what you would consider a typical or standard die to look like. The Low variety uses less inserts and lower grade materials. The High variety uses more inserts and higher grade materials to sustain the tool under high force and high hits conditions.

The quoting methods in the stamping industry are a blend of black magic and bullshit.

You have your good old fashioned Sight Method. That is where someone looks at the part and says, “That is a $250,000 die right there.” How the fuck do they know that? “Experience.”

Then, there is the Length-of-Line Method. You know that one. Measure some trim lines, a few form break lines, then multiply some random number to the length to get your quote. Now, if only you were selling line measurement services instead of stamping dies …

Next up on the list of fancy quoting tools is The Matrix. This works alot like an Ouiji Board. Find the row that has your blank area dimension, then find the column that has the number of stations. It points you to the price. You can almost hear some dead estimator whisper the price to you from the die graveyard.

My favorite is the BOM Method. This is the Bill-of-Materials-thing-a-ma-jig that bases the quote on a standard stocklist of items found in a tool. A guide pin here. A spring there. Keep in mind that the component prices are someone’s best guess and not real numbers. The theory is if you know the component costs, you can creatively guess at the cost of the tool. Somehow.

Here is where all these methods suck: The customer says your price is 30% too high. How could they know that? Using the above methods, you don’t even know if you are high, low, or somewhere in between.

Wouldn’t it be nice to go back to the customer with a rational breakdown to show them that your price is 5% above cost and there is no room to take out another 25% if they buy it at cost?

I would really like to see companies that are looking for a quote to build a die to stop asking for dollars and cents and start asking for hours and pounds.

Why?

Because the cost drivers in any die are machine hours and material pounds. It is that simple. Everything else has a direct relationship to machine time and material weight.

Estimate the 2D and 3D machine time (including wire and sink EDM, of course) and the amount of material required to get the job done, and you will be able to determine the real cost and prove that cost to the folks that think they can put the “your prices are too high” gun to your head.

There has been much discussion on the web lately about quoting and costing software for stamping dies.

Costing is not quoting.

The calculated or estimated cost has no practical relation to the quoted price for two reasons.

First, the price that will win the work relies exclusively on what the customer is willing to pay. End of story.

The customer does not care that it will cost the die shop $500,000 USD to build if they are willing to pay only $400,000 USD for the job.

Even if your costing software is accurate to +/-0.005%, a cost of anything over $400,000 USD is a loser.

Second, most die shops have no real clue as to what their geniune costs are. At the quoting level, it is guesswork.

At the historical level, the data is usually poorly tracked, collected, and reconciled. The historical data should be the baseline for future work. Without this important data, all decisions moving forward are simply guesses.

Again, even if the historical data was valid and the costing method – software or otherwise – was accurate, the only thing that matters is quoting the maximum that the customer is willing to pay and work like the devil to reduce costs once the quote becomes an order.

There was a discussion on a LinkedIn tool and die group recently about how long it takes to quote a progressive die.

One response was simply “two weeks”.

This is wrong. Even assuming it takes two weeks to quote each progressive die, the statement “two weeks”, to be accurate, implies the die shop has 100% conversion.

In other words, the assumption is they convert ALL quotes to sales.

On average, die shops earn one sale for every three quotes.

The quoting math, or qCalculus, then, is two weeks x three quotes = six weeks.

The die shop has to recoup this cost of revenue. How much is it?

Alot.

Assuming the shop burden rate is an average of $65 an hour per person with one person working 40 hours a week to quote, the cost of revenue is 40 hours / week x 6 weeks x $65 / hour, or $15,600.

To put this in perspective, it usually takes four weeks to design the die. It takes 50% longer to just quote one sale.

But wait! It gets worse.

If we assume the “average” progressive die is somewhere between $50,000 and $55,000 to design and build NET of quoting costs, then the cost of revenue is roughly 30%.

In order to maintain a modest profit margin on the job, a $52,000 die build project needs at least $15,600 in engineering changes for the shop to break even.

Yes, break even.

Remember: quoting is not selling.

Here is a two minute video demo of Speedraft technology. The use of mantras is evident here: We eliminate RFQs. This is taken to the next level with ” waiting is not buying” and “quoting is not selling”.

httpv://www.youtube.com/watch?v=5qK8ZdvX54M

Clear. Concise. Compelling.

I was involved with a formability analysis project recently. Here are the facts:

1. The automotive OEM had performed incremental simulations on eight stampings;
2. The Tier One die shop sent out three RFQ packages to three Tier Two die shops;
3. The Tier One die shop requested one-step simulations from each Tier Two shop.

This makes no sense. It is economically irrational. Why have three companies replicate each other’s work?

And why settle for lesser quality results from a one-step when the OEM has full blown incremental simulations available for quoting purposes?

It is a quote, not rocket science. Why spend the time and money doing something inferior to something that already exists that closer to reality? Not once, not twice, but three times?

Even the Tier Two company said “you can’t fix stupid”.

httpv://www.youtube.com/watch?v=0gxKStPXyn8

Like Ron White says, “stupid is forever”.