FDM Hubs, I need your input

[Redwood]

Hi @twiesner, awesome response to the question! It always great to get feedback on ideas here on Talk. We are currently working on an Infill article for the knowledge base but if you have any suggestions of things you would like to have included (other than what you have talked about above) please let me know. Cheers, Ben

[Perry_1]

The real question here : what actually caused the bad review? in quite a few cases we did objectively see good reasons for these objections- how?

Or was the communication with the end customer poor?

Or was the Hub not good at it?

Do you have examples where experience and communication was good and strong, but it did not prevent the customer issue?

Lets take the example of poor tolerances. Key here is, did the hub know there would be a tolerance issue? Did they communicate that to the customer? If the customer still paid, then it would be disingenuous for the user to give a bad review.

Nothing new in this. Print contains part A, which is supposed to fit into part B. Customer is not aware that will not happen without some corrections on tolerances. If HUB knows it, and explains it, then the customer is at fault.

Now, lets take an example where there is no part B. If it is a part where the HUB should have mentioned it, then its on the hub. I have had a few of these, where I have had to ask “what is this for?” or how close does this need to be?

Point is, on these ones you looked at, were there good or poor communications?

In situations where the customer complained about the price of the support material, after paying extra for the support material and removal, I am not sure there should be a legitimate customer complaint. And if 3dhubs can figure out how to pre-price support without a human seeing the model, that would be some fancy software.

Look at hubs with a lot of prints and good reviews. Count the average words exchanged before the order is accepted/paid. You might have evidence of what works, as opposed to trying to fix the unfixable.

I think, the more you try to automate up front, as opposed to demanding good communication between hubs and customers, the more poor reviews you might see. Once you set a tolerance that is acceptable, many customers who would have been thrilled with their raspberry pi case, would suddenly see a reason to be unhappy…

[Enza3D]

On the topic of support, something needs to be done here to make the customer aware of this. It is very hard, if not impossible, to pre-build cost for support into the price of printing for the initial quote the customer gets. If I try to build the cost in, my prices inflate significantly and make my Hub far overpriced. If I don’t, I get complaints from customers about having to pay for support.

It would be very difficult to try and incorporate this into a cost quote, but something as simple as a pop-up informing the customer that support costs will most likely be charged and are not included in the quote here (like you would do with shipping) would at least prepare a customer for this. Supports are an inconvenient necessity of 3D printing, and should be treated as such.

[Lembach3D]

Hello Enza3D

i agree with you.

but maybe this can even be extended to a general notice to the customer, that there might be additional costs. Now with the new delivery cost Setup, most of the times the customer knows the final Price right away. which is great.

but sometimes the part is too big, Needs to be split into more pieces, Needs Support, Needs other kinds of adjustments to be printable. This is impossible to include automatically. But i think your idea of a notice, that the shown Price is stiil an estimate is good.

I also get many orders from “first time hub users” when you see that number, you know already you Need to take some more time to explain. But the time is absolutely worth it, because if it goes well, the customer will most probably come back for another order later.

With so many new users, i think the Ratio of complains is really great. once they got some experience the order process will go smoother. the Ratio will get better over time…

[Perry_1]

@Enza3D Couldn’t agree more!

I wonder if the software could look at the overhangs, and suggest “This model may require support material, which may or may not incur additional charges.”

Or, for every customer->

“Some models require support material, additional modeling, and other work which may incur additional costs.” or something like that.

[Enza3D]

Hubs already has to implement some kind of slicing software to generate the print estimate, so it should be able to identify features like overhangs somewhere.

A general “memo” to the customer when a print is uploaded that the quote generated is just that, a quote, and not reflective of the final price would be fantastic.

[twiesner]

Had no idea this existed either. One factor though, majority of customers don’t want to read extensive amounts of text so many will skim or not read said documents as 3d printing may be a one time deal for them.

[cobnut]

@Enza3D I guess there might be an issue if the print isn’t correctly orientated when uploaded. I’ve had a couple of orders where the STL is “upside down” or on its side. I don’t know, but it might be impossible for software to check orientations - maybe not, I don’t know how smart it is.

[cobnut]

Well, this can’t be a coincidence…

3D Hubs acquires Printivate to improve 3D file checking, repair &...

[Enza3D]

This is very true and always a possibility. That’s why I think a blanket prompt stating that the “quote is not reflective of the final cost, and does not include support structure, etc” will be the most effective (and least limiting) way to go about addressing this.

[cobnut]

You’d have to be very careful with statements like that though; they’re likely to scare customers away. We all want to know exactly how much something is going to cost us - it’s a principle of good sales marketing - so any message suggesting the initial price may change would have to be worded very carefully.

For a start, ideally, there should be no message at all unless the Hubs software can detect a likely overhang. If there is a risk of support requirements, I’d suggest the message should be more along the lines of “Please discuss the printability of your model with your chosen Hub” and leave it at that, not mentioning any potential price change. If the Hub decides they need support and wants to charge extra for it, that’s then a discussion between themselves rather than being something that instantly says “you’re likely to get additional charges”.

[Enza3D]

That is true; the majority of companies I do prototyping and machining work with just explicitly state “Price as shown here is a quote and is not necessarily reflective of the final price”. That’s just typical invoicing practice in the engineering industry, so I’m used to it.

I think the biggest issue is that a lot of customers (especially new ones) come in with the expectation that the price they are shown on the main Hub search page is the final cost, when that’s really not true, but I can’t blame the customers because that’s not really made explicitly clear anywhere. The software Hubs uses to calculate the volume isn’t even right most of the time (I usually have to adjust it manually to match what my slicer gives me). Leaving out anything about price changes is a bit deceptive (I think), but saying “This is an automated quote and may not reflect final price” should be enough to get the reality of that quote across without scaring off too many customers.

It’s important to distinguish the difference between a site like this, and one like Amazon or eBay. On those sites everything has hard set values and the seller has the final product in their hands, so you can know exactly how much you are going to pay. Hubs is not like that as you well know, and it’s not even companies like Shapeways because Shapeways has one set “operator” and can quote exactly because of that or they sell from shops like Etsy. Aside from shipping cost and raw material costs, there are so many unknowns at play here that it would be impossible to get an exact “You will pay XX for this” here.

[Filemon]

We’re still testing the first few articles and will ‘officially’ launch in a few months

[Robin3D]

Hey guys!

Thank you for all the responses with 48 hours after placing the thread! Already some really good insights on what you think are the main causes of any issues customers have with the prints they receive.

First of all I will elaborate on my sentence that stated: “in quite a few cases we did objectively see good reasons for these objections.” I left this quite vague on purpose to see what you guys would think are the main things that should be in the guidelines for FDM printing. We noticed that only a small portion of the orders with complaints had to do with the fact that the expectations of the customer were unrealistic. Most of the customers had a legitimate complaint on how the final print turned out.

We had a look at the orders where customers indicated they were not happy with the prints and by judging the pictures posted on the order pages we noticed the following recurring issues:

Surface quality, dimensional accuracy, warping and poor support removal.

Some first thoughts on what topics should be part of the possible guidelines, please note this is all very much up to debate and the whole reason why I made this Talk thread to begin with.

• Surface quality: This one is a bit tricky but very important when talking about FDM quality. Obviously, the customer should be expecting visible layers in FDM printing but what about layer consistency, the look of parts that have 45 degree angles. The fact that some printers are able to print more consistently than others and some will leave marks where others are able to print a very smooth surface.

• Dimensional accuracy: This was another one of the issues that came out of our analysis. Quite a few customers were not satisfied with the dimensional accuracy on the prints they received. How can we prevent this from happening in the future and what is a reasonable percentage the final print is allowed to deviate from the design?

• Support & Support Removal: I already picked up from the comments in this thread that you guys definitely understand the issues with automatic support calculation (although we do have a dedicated team looking into ways we can improve this). For now we cannot calculate support material on FDM prints and we will have to rely on our Hubs to add these costs during the order process. I would like to propose to make it standard that every Hub that adds support material to orders also includes removal of supports in the pricing.

• Communication: As a couple of your already pointed out, a large factor that contributes to the success of an order is good communication between Hub and customer to set the right expectations on what the final result should look like. When handling orders, make sure that you and your customer are on the same page before starting the print.
  When the print it finished, it can be highly valuable and time saving to place photographs of the print on the order page so your customer can have a look before shipping the parts.

@Lembach3D, I’m not the best writer so I don’t think I’m quite ready to start writing a book about FDM printing but who knows, maybe somewhere in the future But I get your point, FDM printing is complex and is best discussed on the order page but still I feel that some guidelines will definitely be helpful for both Hubs and their customers

@Perry_1, Interesting thought to see if there is a ratio between the amount of communication on the order page and the chance of the order going into the Dispute state.

@Enza3D, I completely agree with the fact that we could give a heads-up that additional costs might be added to the order for support material because 3D Hubs cannot calculate this upfront. I’ll communicate this to our product team to see if this can be added.

Thanks again guys, I look forward to continuing the discussion.

Best,
Robin - 3D Hubs

[Enza3D]

@Robin3D

In regards to dimensional accuracy, that is a very tricky one with FDM. Much like surface finish, this will vary from printer to printer depending primarily on the nozzle size.

The bigger issue, I think, is that for a good final print tolerances need to be built into the design of the model from the start and most of the time, that does not happen. The reality of 3D printing (or any manufacturing technology) is if you have a hole that needs to accommodate a 2" diameter pipe, that hole can not be 2" in diameter as well. The printer will never exactly replicate the model, and it’s unrealistic to expect that. These are core concepts of mechanical design, but not something your average customer will understand. Even if they did build tolerance into a model, the right tolerance for one printer does not guarantee it’s the right tolerance for another. I know with mine ±0.4 mm is a realistic tolerance for loose fit and I can push ±0.2 mm for snap fits and gears, but that’s only for me. I could buy a new printer and it behaves very differently. This is also why I make it explicitly clear to my customers that I will not guarantee the final part fits dimensions they specify unless I design the part, I only guarantee it will match the model to the best of my printer’s ability.

I think unfortunately, the reality is you will never get true “dimensional” accuracy from a setup like 3D Hubs, unless the specific Hub models the part. There are just too many factors that are not controllable at play here and it’s not like a standard prototyping shop where you get toleranced drawings to work from. A good Hub can definitely get it close to desired if they work with/can get information the customer and the Hub is willing to correct models/the customer is willing to pay for remodeling, if necessary. I do not think it is fair to expect a Hub to print a pre-made model to exactly fit unknown requirements (especially from an STL, not a solid model), the only thing that should be expected is the print matches the uploaded model as closely as physically possible. I also do not think it is right for 3D Hubs to promise “dimensional accuracy” to the customer, when there is really no way to guarantee that.

It’s like if I order some graphic to be printed on a t-shirt from a silk screening company; it is not the shop’s responsibility to ensure that spelling, design colors, and alignment is correct. Their only responsibility is to replicate the graphic as best as they can on the t-shirt in a location range that I specify. It is on me, the customer, to make sure the design is correct in the first place.

The other reality is most customers here are not trained in modeling/designing for manufacturing, nor do they have any interest in learning (obviously, this is not true of all customers). They just want it printed so it works how they want it to, which is not an unreasonable demand, but the understanding has to be there that this is not a perfect process with even the best operator and printer, and there is a significant increase in scope of the order to make sure a part fits dimensions.

[3DPrintandStuff]

Hi Robin,

In regard to Communication.
I believe it’s lack of it from the customer and 3D Hubs about the process of “3D Printing” by the fact that term covers many diffident machines and technologies and some produce better results than others and have higher costs, which is why they produce better prints.
I have had a hub on here since the early days, be though the changes bad and good, worked though problems with FDM printing upon my machine, tried to educate users that what you are asking me to print is going to take a while and results might poor. This upsets me and I have scraped prints and printed another before sending to the customer. This is something that 3D Hubs states I have to ask the customer for.
Currently my hub has completed 21 orders but only had 14 reviews. My last order was 475 items in PVA, 20x 20 x 40mm pods & 20mm high pods with 0.8 walls and lids some with/without holes or “as thin as possible” as the given brief. I was able to process a small order from the same customer before this one within my 3 day deadline, this one I was given a review for. Customer was impressed.
With the 475 item order of course it was not possible to complete in that in 3 days, this was communicated to the customer and the deadline date extended until I was happy that it would produce usable prints for them to use.
The order was accepted and the amount was paid to 3DHubs, so printing commenced, in the end I broke this one order down into about ~12+ print runs ranging in time under 5hrs to over 11hrs, also bought a new machine and set it up. Took about 3 weeks to complete and used just over 500grams of PVA filament, by the way ESUN PVA rocks!!! prints fast and takes the heat!

During this time despite my questions upon the order the only response I got was at the start was “Just let me know when it can be delivered” Also had the delay on the release of payment from 3D Hubs, which I understand the need for but customer has received the prints (i know that, as signed for), they have paid but won’t log back in to click a button saying “it’s all fine” and the Hub has to wait up to 10 days for the payment from 3Dhubs to arrive, which causes cash flow problems.
I would like to hear or see and understand why I have 20 completed orders but only 14 reviews. What was the reason for no review when I worked hard to produce the prints, even scraping some because of failures. Check out my instagram account and my hub order update for further details and photos. I’m proud that I was able to complete the large order. I just wish the customer would understand the work which was done to produce it. I’m thinking the customer wasn’t happy and felt not able or wanted to leave a negative or not quite positive experience. I just don’t know.

Maybe have a badge called “Blood, sweat and tears, this hub goes beyond the call of duty with their order” or I just add another 20% to account for the time it takes to get paid for the order from the time the customer gets the print, which will most likely end in not getting any orders, it’s a fine line!!

Many thanks, keep up the good work

Chris

[3dprintingguy]

I would love to hear how hubs out there price support removal, are you doing it for free? I have to factor in human wages vs machine running costs because minimum wage in Australia is $17/hr And support removal sometimes isn’t an easy task! I figure clients will be better 3d designers when they need to think about supports, they learn quickly after they have to remove supports once or twice

I don’t just dump it on them tho, I link them to 4-5 articles and 5-6 you tube videos and also offer to do it at a cost

[Robin3D]

Thanks for that Chris, very impressive order. I checked it out, that’s a looot of pieces! After every order we do invite customers by e-mail to leave a review for the order and they get a reminder automatically if they do not. If you want I could send a separate message to your customer as well but in some cases, customers simply don’t leave review.

Having a badge called “Blood, sweat and tears, this hub goes beyond the call of duty with their order” is actually not a bad suggestion. The execution would be a bit different but another way of standing out for Hubs that are able to print very high quality FDM prints besides just completing the Marvin review might be another way to go. What do you think?

[Enza3D]

I think specialized badges are a great idea, and something I’ve talked about with others before. Maybe figure out some way to develop certifications for things like high-quality FDM, engineering design, etc. and then the Hub can have a badge certifying they are in fact what they say they are.

I know as a customer, I’d be much more comfortable prototyping with a Hub “certified” for engineering design than with one that is not, and I know I’d be working with people who know what they are doing. I’m not exactly sure how you’d go about doing this, but I think it would be a huge benefit to the community and give a lot more validity to what a Hub can claim. Really just a formal verification of what a Hub already calls their “specialties”.

[3DPrintandStuff]

Thanks for checking out that order. It broke in the new printer as I was printing as many of them as I could fit upon the printer’s bed each time.
I agree that most people won’t leave a review as I spoke to others about it. The response I got back was I might not/don’t like to leave a review which is not as good as it was/could be, sort of middle of the road but don’t wish to upset them. This has been brought on by need to have 5 stars upon each order with no understanding that the printer might have an off day or a print fail

I have also thought the method of pricing the prints using volume makes it hard and expensive when to comes to large item numbers of each and large volume prints. If the pricing was done by the meter it would be cheaper (better) in the view of the customer but for the hub (not as good) but it would allow the hub to add human hours costs, electrical costs etc to that meter price but still come under the volume print price for the same item. With that large order the start up cost paid for a 750gram reel of Orbi Tech PVA @ £75 but didn’t allow for fails. problems with filament not working, the need to man (be present when printing) the printer. When I spoke about the order to others they sort of mocked it by saying that it’s not possible to print that many items, 3D printing will not do large manufacturing runs of single designs, as it’s not as rapid.

The above might explain why people don’t wish to leave a review. Also having an average price paid for an order within each area. So that the hubs can they see if under or the over the average. As in my area one hub has dropped the price of their prints so low that the other hubs miss out on orders. Sort of Is the hub running a business or as a hobby?