Xometry Premium Partner

[Marvel]

While @nateacox and I have similar experience and success with Xometry, I don't compare to his size. I would fall into the category of his comment "please don't message me" Although I don't have a Tormach or a 1998 VF2 - but don't worry @nateacox I won't be reaching out for any help, I don't need it. I am however about 1/10th the size of his operation. I have 4 CNC machines, a VF2SS, VF3SS, EC400PP, SL20 and I've done almost exactly 1/3 of what he has done with Xometry in jobs and sales, any given year Xometry has amounted to 50-75% of my customer base, and that's dependent on how busy my other customers keep me. I do view Xometry as another customer, not just a sourcing company, but I have more control over the work flow, whereas some of my customer just send PO's without a heads up.

 

[Radar987]

I don't have anything against Xometry, but for those of you who are heavily embedded in it, have you thought about what you would do if Xometry went kaput?

Check out Xometry's stock. It's down 80% since IPO.

 

[boosted]

I don't have anything against Xometry, but for those of you who are heavily embedded in it, have you thought about what you would do if Xometry went kaput?

Often. This concern prevents me from getting more heavily invested with them. We are specifically focused on high mix, low volume, complex geometries, and rapid prototyping. On paper Xom is a perfect fit for us.

My experience is similar to Nate's; though not quite at that scale. 7 man shop. 3 five axis with automation, and one vertical under about 6500sqft. We are also in a super expensive place to do business relative to the midwest.

It has taken a long time to build relationships, learn how to play the game, etc... But Xom is a profitable customer for us, and they come with some major perks like the ability to ignore them when we are busy with other customers, and zero drama payment that always comes on time.


I have had many people ask what the secret to making money with Xometry is, and there is no single magic thing. We are ultra-premium partners, ITAR, AS9100D, and CMM certified. All of the extra tiers may increase the volume (and certainly complexity) of work available on the board, but I don't know that any of them really increased profitability. There are tons and tons of simple jobs on the board that are being done by desperate shops and novices at cost. In order to make money, you must take the jobs that need an expedited form tool, have weird geometry, require 30 hours of programming in 2 days, or some other BS. You also have to counter-offer pricing for those jobs if you want to get paid. It has to get sourced - if you are the only bidder, they will most likely meet your required cost.

I tell anyone who will listen that we make more money in the office than on the machines. It is very typical for us to spend a week programming a part that only runs for a day. There is a chaotic dance of staging everything to get a pile of complex expedites through every week. The way I do it is not for the light hearted, but with the right resources in place it certainly can be done!

 

[F35Machinist]

You are 100% right, you do not have to accept it or be party to it.

Xometry's response time is typically terrible, I can agree to that! I don't see very many 2 day lead time jobs, I typically see 3-4 weeks, are there shorter lead times, yes, with terrible pricing, of course there is.

I don't take very many low paying jobs, so a $600 job I would more than likely pass on it, occasionally I do, when I'm bored and have an open machine and my other machine has a long cycle time, and if I have material and tooling in stock I may grab a few jobs under $600 that I can bang out in a day, cause to me if my second machine is sitting and my other has a 3-4 hour cycle time and I can bang out 2-3 $4-500 jobs in a day, that's an easy $1500.

I wasn't necessarily referring to your exact $600 job, it was more of a in general comparison to shops stocking certain tools that a lot don't and/or having a lot of a specific tools some don't. Another example would be insert tools, I have just about every insert tool from tangless and tanged along with every STI thread gage, including Keensert install tools, if someone was to take a job with multiple, they get expensive real fast, I've had jobs that require 6 different insert sizes, the insert tools alone would come close to $1000 now add in STI thread gages you could be pushing $1500-2000 fast for one job.

I don't know about different tolerances, I've never come across that. But there's a tier system, I've heard things are different for newer Partners in recent years, could be a major factor. But there are a lot of large medical and aerospace companies that run tier systems no different, when you first become a vendor you are at the bottom, you need to prove your shop and capabilities. No company wants to give a random shop a $20k PO to find out 1-2 weeks later they can't even accomplish the job, need to prove what you are capable of, any shop can have the capability but doesn't mean they have the experience.

I acquired a new customer, locally, last December. First meeting they told me, the first few months I would get the bottom of the barrel work, with short lead times and low paying but after so many jobs of proving my capabilities, quality and being able to hit due dates I would move up tiers, in about 45 days they quit having me quote work all together and started sending me PO's with $0 and pre determined due dates, with the understanding I would bill T&M and they would cover any expediting cost, and they informed if I took advantage of it I would go back to quoting jobs and competing against other shops that are quoting.

Also people need to realize, not every machine shop has the same overhead, not every shop needs to clear $150-200/hr, or whatever it may be. I am a home shop with a VF2SS, VF3SS, EC400PP and SL20, I have very little overhead and can be very competitive if needed. If I became desperate for work, I could easily quote at $50-75/hr and be perfectly well off.

If a customer told me they were going to send me shit work for no money for months as a test, I would just tell them to go somewhere else. Maybe I should be referring them to you Marvel?

It's true that brand new customers don't often start with big POs, but in my experience it gets there pretty fast if it's going to get there at all. Maybe 5k order for a few units to test you out and then the real order after they know you can do it. There's no running the gauntlet with months of trash work.

Any time I get a new inquiry that sounds like they're shopping around I tell them that they should look at rapiddirect and pcbway for lower pricing. There is no point in trying to please price shoppers that will ditch you when they hear about rapiddirect anyway.

 

[Marvel]

I don't have anything against Xometry, but for those of you who are heavily embedded in it, have you thought about what you would do if Xometry went kaput?

Check out Xometry's stock. It's down 80% since IPO.

View attachment 395439

Any customer can go “kaput”, happens all the time. I worked for a large machine shop about 10 years ago that was heavily invested in two medical device companies. I don’t recall exactly what happen but the two medical device slowed down or started using other vendors and the shop took a massive hit out of nowhere.

But my advice is you never depend on one source of income in business, you should have multiple customers.

I can’t say for certain that Xometry is doing this but a lot of business run on NOL for tax purposes and a publicly traded company on paper like this will show what you see. I own a storefront business as well and it’s exactly what we did, you carry forward losses from previous years to lower your tax liability.

 

[Marvel]

If a customer told me they were going to send me shit work for no money for months as a test, I would just tell them to go somewhere else. Maybe I should be referring them to you Marvel?

It's true that brand new customers don't often start with big POs, but in my experience it gets there pretty fast if it's going to get there at all. Maybe 5k order for a few units to test you out and then the real order after they know you can do it. There's no running the gauntlet with months of trash work.

Any time I get a new inquiry that sounds like they're shopping around I tell them that they should look at rapiddirect and pcbway for lower pricing. There is no point in trying to please price shoppers that will ditch you when they hear about rapiddirect anyway.

I didn’t say “shit work for no money” I said “bottom of the barrel work, with short lead times and low paying”

In the first 45 days, I was easily averaging $75-$100/hr for this customer and once they started sending me work I wasn’t quoting, I understood what they meant by “bottom of the barrel” it wasn’t bad work at all, it was just fairly simple no brainer parts, +/-.005” typically were 6061 or Acetal with two week lead times. I still see some of that work, but I also get more complex parts, different materials, different types of stainless, titanium, parts with multiple process finishes. I ship on their UPS account or I feel the parts are too big to ship, they send their own driver to grab them.

 

[CLW]

We have had a great experience with Xom as well. Everything Marvel, Boosted, and Nate have said is 100% truth. We have added 2 new Hermle C400's in the past 6 months, without Xometry we would'nt have added those machines. The prices do fluctuate a ton, and right now in my opinion they are lower than the norm. There are always Jobs on my board that are due tomorrow and the material cost more than the Job pays. If you are looking for 3 axis work that you can throw in a vise then you might struggle to make money with them.

 

[Marvel]

Often. This concern prevents me from getting more heavily invested with them. We are specifically focused on high mix, low volume, complex geometries, and rapid prototyping. On paper Xom is a perfect fit for us.

My experience is similar to Nate's; though not quite at that scale. 7 man shop. 3 five axis with automation, and one vertical under about 6500sqft. We are also in a super expensive place to do business relative to the midwest.

It has taken a long time to build relationships, learn how to play the game, etc... But Xom is a profitable customer for us, and they come with some major perks like the ability to ignore them when we are busy with other customers, and zero drama payment that always comes on time.


I have had many people ask what the secret to making money with Xometry is, and there is no single magic thing. We are ultra-premium partners, ITAR, AS9100D, and CMM certified. All of the extra tiers may increase the volume (and certainly complexity) of work available on the board, but I don't know that any of them really increased profitability. There are tons and tons of simple jobs on the board that are being done by desperate shops and novices at cost. In order to make money, you must take the jobs that need an expedited form tool, have weird geometry, require 30 hours of programming in 2 days, or some other BS. You also have to counter-offer pricing for those jobs if you want to get paid. It has to get sourced - if you are the only bidder, they will most likely meet your required cost.

I tell anyone who will listen that we make more money in the office than on the machines. It is very typical for us to spend a week programming a part that only runs for a day. There is a chaotic dance of staging everything to get a pile of complex expedites through every week. The way I do it is not for the light hearted, but with the right resources in place it certainly can be done!

I'm quite curious what CAM system you are using and of course the part that was programmed that took 30 hours! I know you probably couldn't show or share the model.

I don't think I've had a part take longer than 8 hours to program, I've had a couple 30+ hour cycle times and programs are pushing 15MB file sizes.

I've spent a significant amount of time dialing in my CAM default settings or what some would call "templates" so I do minimal tool selecting and changing of feeds and speeds. Everything is based on stock material and operation to pull in all my predetermined data. It took me about 8 years to actually invest some time into perfecting these settings but it didn't take long to notice how much faster I was producing programs.

 

[boosted]

I'm quite curious what CAM system you are using and of course the part that was programmed that took 30 hours! I know you probably couldn't show or share the model.

We use hyperMILL over here. The vast majority of our work is ITAR or some type of NDA, but here's a couple pics of some BMW oil pans. I programmed these myself. It took about 80 hours. Lot of tricky undercuts and porting that aren't fully visible in the pictures.

I have invested a ton of time into automation on the machine side, as well as programming. Once you get deep into five axis though, part complexity (and programming complexity) can go up exponentially. I have done several parts that have over 100 tools. No matter how good you are - that's gonna take some time!

Most of the Xom jobs we make money on have a massive programming burden, and other five axis shops are either at capacity or can't move fast enough to hit the due date.

 

[boosted]

I don't think I've had a part take longer than 8 hours to program, I've had a couple 30+ hour cycle times and programs are pushing 15MB file sizes.

Again, that's the difference between 3ax vs 5ax worlds. I am working on a random, nothing special part on the DMU50 this weekend. Program size is 53MB. ~ 8hrs each to run @ a quantity of 12. Took the machinist about 60hrs to program. A big part of that programming time commitment is the fixture design.

 

[Garwood]

Those oil pans are beautiful!

 

[the average Joe]

yeah those deep side cuts look amazing

 

[the average Joe]

I am the owner of a shop with 10 employees, 20 CNC machines, a workspace of 15,000 square feet, and we operate on a single shift. Our basic hourly rate for simple work on inexpensive materials is $100, and this rate can increase depending on the complexity of the job. Compared to businesses located in larger metropolitan areas or California, our operating costs are relatively lower.

Occasionally, we receive jobs from customers like Xom that generate lower profits. Generally, we take on these jobs during periods with available capacity, or to build goodwill with the customer, or even to secure multi-year contracts. We currently have a job for a local customer involving 400 pieces that runs on a 5-axis mill, paying us only $67 per hour. While the hourly rate is lower, the job runs for 16 hours per day, 6 days per week with minimal involvement from a machinist, and minimal tool usage. Additionally, the customer supplies the raw material and performs all secondary operations in-house, making this job a win for us, as it provides high utilization rates.

Regardless of the customer, working in this industry requires paying to play. Every hour spent interacting with customers, quoting jobs, updating projects, and waiting for responses is a cost. Paying for leads or a salesperson, reaching out to potential customers, providing tours of our shop, and even providing entertainment or prizes to customers are all costs associated with doing business. Furthermore, waiting to receive payments from customers is also a cost. Therefore, working with Xom incurs costs, just like working with any other customer, and I believe that the value gained from these costs justifies the expense.

I understand my job board looks very different then the majority of the people on here. But it didn't just magically start like that. I paid my dues, and did my time to cultivate the relationship with them. If you aren't interested in doing that, then don't.

If however you have a serious shop, legit equipment, and a track record of doing good work, feel free to PM me. I can put you in touch with the right people there to help get you aimed in the right direction with them. If you only have a tormach, or a 1998 VF2 and are a one man shop, please don't message me. If you don't have a good selection of measuring tools, or a quality manual then don't PM me.

Now you're making fun of my '98 vf-2? I'm feeling called out right now : )

 

[nateacox]

Everyone has to start somewhere. I started with a 95 VF2.

Stock prices aren't the best barometer for business success. If they went under, I would pick up the pieces, and use the same principals and find another customer.

 

[the average Joe]

We still have our '94 VF-4 too, complete with the old CRT orange screen. Probably the least troublesome HAAS we've owned.

*knocks on wood*

 

[alek95]

I wish I had a single old Haas to start my own shop.

Unfortunately the work visa that lets me work in the states doesn't allow me to start my own business, and I do not wish to return to Canada.

So everyone who has the opportunity to own an old Haas should count their blessings, in my point of view.

 

[the average Joe]

The grass is greener where you water it.

I honestly like all our old Haas machines. They get the job done, are still fairly accurate and once you start fixing them yourself repair costs aren't too bad.

We have an older HAAS VF-8 that keeps us really busy, to the point that we're considering a more deliberate strategic move towards acquiring at least another large machine and some other support equipment to tackle more large format/feature rich stuff. Apparently, not many shops want to do that kind of stuff.

Even if you can't start your own business, can you buy a machine and do some side work? I have a side gig doing medical device design and I just run it as a private contractor.

 

[alek95]

The grass is greener where you water it.

I honestly like all our old Haas machines. They get the job done, are still fairly accurate and once you start fixing them yourself repair costs aren't too bad.

We have an older HAAS VF-8 that keeps us really busy, to the point that we're considering a more deliberate strategic move towards acquiring at least another large machine and some other support equipment to tackle more large format/feature rich stuff. Apparently, not many shops want to do that kind of stuff.

Even if you can't start your own business, can you buy a machine and do some side work? I have a side gig doing medical device design and I just run it as a private contractor.

My work visa expressly prohibits me from any side work, even driving for Uber.

Hopefully I can find an employer that is willing to get me a green card. That would allow me to start a business. I work evening shift monday to thursday. That means I am free until the afternoon five days a week, while still keeping my well paying day job.

 

[Marvel]

Again, that's the difference between 3ax vs 5ax worlds. I am working on a random, nothing special part on the DMU50 this weekend. Program size is 53MB. ~ 8hrs each to run @ a quantity of 12. Took the machinist about 60hrs to program. A big part of that programming time commitment is the fixture design.

That makes a lot more sense. I have minimal 5ax experience, but have done quite a bit of 4th axis with simultaneous operations and I know where the programming time starts to add up fast especially when you have fixturing. I recently had a job that ended up being a 7 hour cycle time with quite a few simultaneous operations and I was committed to designing and building a fixture to get one part out of the machine complete per cycle so it allowed me to run 3 parts in a 24 hour period, I timed it to load the 3rd part and run overnight, I had just as much time, maybe more into designing the fixture and programming it to ensure repeatability.

When you start getting into rotating, even if its just pre position, but even more so simultaneous you have a lot more to pay attention to with clearances, tool holders, fixtures, all that stuff actually matters whereas in a 3ax world you can damn near get away with paying minimal attention to any of that.

 

[the average Joe]

I can't imagine crashing a 5 axis machine is a cheap fix.