Possible Stoppage of the BatchPCB service

Archive BatchPCB
1

Hi,

We’re pretty strongly considering stopping the BatchPCB service [aka, finish all currently in progress orders + accept no new ones].

As some of you probably know, we originally started batchpcb as a way for hobbyists to be able to get small qty 1-3 boards made cheaply, at the expense of long and flexible leadtimes.

Its not a service we make much/any money off of, and in recent times, the workload has just become to great for us to maintain the current status quo, as more people start to use it for time sensitive projects, which is not the service model we’re geared for.

Does anyone have strong opinions for keeping it open [or for shutting it down, for that matter]?

Cheers,

–David Carne

2

Although I’ve only used the service once, I like it, and am planning on using it again. For the first spin of a design, I’d rather pay ~$25 for one coaster (er, board…) than $60 or more for a larger quantity of something that might not work.

But, you guys are a business, and the bottom line is what matters. If BatchPCB is breaking even, it might be justified as a service to the hobbyist community, but if it is a significant resource sink, your decision to shut it down would be understandable.

Here are some idea that might support hobbyist use while discouraging commercial customers:

  • Limit to 2 or 3 boards per order, one order per week

  • Limit total area of order to something small (16 in^sq, maybe)

  • Quote two month lead time in large, bold letters, but explain that you’ll make an effort to keep ~21 day turn, maybe with price reduction (or store credit) if two month timeframe isn’t met. Any sort of company that relies on that is crazy.

  • Add “NONCOMMERCIAL USE ONLY” to top and bottom silkscreens

I don’t know about everyone else, but I’d be happy with any of these limitations, if it meant I could still get cheap PCBs for my own projects.

3

I think we might start quoting 1.5 month worst case lead time, 20 days average. While we make the 20 days on most of our orders by far, the 1.5 months will give us slack in the case things do go wrong.

That may prove to be a workable solution for us. Keep your eyes open for a change on the site tomorrow-ish.

Cheers,

–David Carne

4

To be completely honest, it appears as though you have a generally very well received and happily used service that really helps out a lot of people in the hobbiest market. In general, it also seems as though your lead times are better than they’re quoted, and are very reasonable for the price.

If people require a board made quicker, the option always exists to use a different (more expensive) fab house that guarantees turn around time.

On the side, I know that you guys have received a couple of complaints from a few folks in recent days. The unfortunate part about that situation is that you probably rarely hear from all of the people who are really satisfied with how well you’ve done, and even a few of them in the thread have quoted their leadtimes as being well under those estimated. In anything you do, even if you do it well, you’re always going to make a mistake every now and again (and, if anything, that’s something we all know really well from /designing/ PCBs!). Maybe occasionally someones order accidentally falls on the floor and it takes some extra time to process – but it’s to be expected, the occasional mistake happens every now and again. It also might be unreasonable to expect that moving offices and taking some of the service offline for a number of days might proceed /without/ the odd hitch or two, of course.

If anything, I would keep things exactly the same, secure in the knowledge you’re doing a pretty good job!, with the exception of perhaps reflecting on some of the comments people have raised to see if it might help you streamline the process a little more in the future.

best,

silic0re

5

busonerd:
Does anyone have strong opinions for keeping it open [or for shutting it down, for that matter]?

Greetings Dave et al.,

I’m a big fan of BatchPCB, in spite of my recent frustration. I’d like it to stay.

That said, I think a fair and equitable position is for SFE to make money by running it as a business (and not a charity), and for users to understand the limitations. There has to be more than ‘cheap and slow’ for it to remain viable.

For my needs the biggest advantage is that I can do a one piece PCB for a hobby budget, with reasonable expectations of the outcome. If the next project uses the same PCB, or uses several of the same design, I can reorder.

Faster delivery would be nice, but I found that I can pace my work and do two or more concurrent projects that fill the twenty day timeframe well. I’m not in a position to use (or be tempted by) the “two for one” offers, or “get five day turns for the ten day price” that fill my email box from other PCB vendors.

What has to get better, if you decide to continue, is that customers are better informed of the work progress.

Changing the timeframe to six weeks and keeping the cost the same will not really help - it doesn’t motivate SFE (no profit incentive) and it won’t take long for users to expect three weeks once it became known that the service can support twenty day (plus or minus) delivery.

(1) Start by doubling the cost per unit area.

(2) Not limiting the size to anything less than 24 sq inches.

Which are the limit of the EAGLE standard license (4.0 x 6.0 inches), and also approximately the size of the popular Eurocard (100 x 160mm).

(3) Offer the same shipping options but without the hidden admin charge.

(4) Provide a satisfaction guarantee that any PCB not delivered within twenty days is free (or, given SFE 100% credit against further PCB Fabs). Customers need to see an incentive imposed on SFE.

My BatchPCB projects have been towards the 20 sq inch size. I can see that many hobby and student projects (for breakout boards) might be closer to two square inches, and this might have been the motivation to start BatchPCB and also add the ten dollar admin fee.

Perhaps a minimum order amount is the solution? It gives SFE the income stream to cover admin costs associated with each order. I see other PCB suppliers doing similar programs to get their desired dollars-per-invoice revenue.

I would use a BatchPCB style service at $100 min per order. That would put the responsibility and incentive on me to get the most value for the minimum billing of a hundred bucks!

In real terms that’s two casual restaurant meals for my wife and I. Or, two fill ups at the pump for my SUV. Or, a single ticket to a live musical event. Or, one more parts order from Digi-Key. All examples of what’s competing with my hobby PCB budget.

Comments Welcome!

6

Please keep it running! I’d like my money stay in Colorado.

7

I need my BatchPCB :frowning:

It’s really the only option for a hobbyist with multiple small designs. I’ve used the service almost monthly for the past year, and it’s saved me tons of money. I can honestly say I wouldn’t have been able to afford most of the designs I’ve done without it, being a poor student :shock: The only alternative is to queue up a bunch of designs every few months and cough up for the array fee at a larger board house. I can handle waiting a month for boards from BatchPCB, but having to order elsewhere every three or four months would be painful. If I made a mistake on one of those boards I would be banging my head against a wall for days :lol:

I think most of chastings’ suggestions are pretty good. Certainly increasing the stated lead time to something prohibitively long for commercial users or time-sensitive projects. Maybe get rid of the “panelized in less than four days”. I’m guessing here, but I think that generates a large number of complaints (at least on the forums).

8

bigglez:
(1) Start by doubling the cost per unit area.

(4) Provide a satisfaction guarantee that any PCB not delivered within twenty days is free (or, given SFE 100% credit against further PCB Fabs). Customers need to see an incentive imposed on SFE.

I think the point is that it’s cheap and slow. If you need guaranteed delivery dates and are willing to pay more, there are plenty of board houses that can probably better accommodate you.

One overseas vendor that I know of will do two 4" x 5.5" 2L boards with mask and screen for $78 + $12 shipping to the USA. But, for a similar total cost, I could get three or four revisions of one design from BatchPCB. Considering how often my designs are finished at the first revision, the latter choice usually ends up being more cost-effective.

9

Keep it running, keep it cheap, and keep it slow.

If panelization time is the bottleneck, double or triple the limits. Most of us won’t care, and the few who do care are the ones who piss and moan on the forums. Good riddance.

I really hope you decide to keep it running; it really is a unique and useful service.

10

chastings:
I think the point is that it’s cheap and slow. If you need guaranteed delivery dates and are willing to pay more, there are plenty of board houses that can probably better accommodate you.

Greetings chastings,

As has been revealed by SFE, the current BatchPCB service is an albatros around their neck:

busonerd:
Its not a service we make much/any money off of, and in recent times, the workload has just become to great for us to maintain the current status quo.

Trying to retain the service as is, or adjust the delivery model will fail.

chastings:
One overseas vendor that I know of will do two 4" x 5.5" 2L boards with mask and screen for $78 + $12 shipping to the USA.

Please post details, or PM me if you prefer. I suspect many BatchPCB users have already tested the waters for an alternative service.

Thanks In Advance!

11

I know what you mean by incentives - but in your case specifically, the order was filed with our fab [on time from our end!], and when we expected the tracking number to come in [about a 1 week window, which is how we’ve negotiated the pricing to be so cheap] we didn’t hear anything.

At that point we started following up with the fab [1 day turnaround on all emails due to the time differential], and it turns out that somewhere in the fab the order got dropped [one we’d already paid for + sent to them]. At which point we said, “fix it then!” and then they started rush building the panel for us, and we expect it here any day. I don’t have tracking specifics on that at the moment though, as I don’t run the service [just write code for the bot on occasion].

The reason the admin fee is “hidden” in the shipping, is because thats the easiest/only way to do it at the moment in terms of the shopping cart we use. For a while though, it should saying +10$ S/H to make it quite obvious its not part of the shipping.

Raising the price would allow us to go to a more local fab + take shorter leadtimes, but for a lot of our customers, the response seems that it would be too expensive. The panelization time is usually not an issue, its just been bad while we’re in the middle of the move. [We’re STILL moving board production equipment + setting up the production area]. We’re currently looking at some possible automatic panelization techniques so we can automate more of the processing stuff, to get designs moving quicker.

Cheers,

–David Carne

12

Bergamot:
If panelization time is the bottleneck, double or triple the limits. Most of us won’t care, and the few who do care are the ones who piss and moan on the forums. Good riddance.

Greetings Berg,

Actually, the main complaints have been that the on-line order tracking doesn’t work. It is seldom up to date, leading to user emails (that are seldom answered on the first try), followed by a surprise when delivery dates are missed. Lost work requires a complete do over and doubles the delay. (If the service was promoted as six weeks turn around, a redo would take three months to complete).

All this points to a lack of financial motivation by SFE to keep the service running well. They collect the CC payments automagically upon submission of the orders. Out of sight out of mind.

I’ve suggested that SFE be rewarded for a good job, done well. Also, if there’s a risk they will loose revenue due to inattention (i.e. a delivery guarentee) tracking the orders and updating the web pages will have greater incentive.

I’m strongly in favour of making SFE rich, because I (and others) value the BatchPCB service.

Comments Welcome!

13

Check my previous post - the delay in your case isn’t due to us missing it… its due to the fab dropping the panel.

Regarding status updates, we’re looking into adding a few datapoints [aka, shipment information received from fab / etc]. Right now the only two datapoints we have are: Panel sent to fab [panelization]+ Panel received from fab, at which point we ship it out the same day it gets to us.

6 week turnaround would not mean a 3 month delay - we’d still strive for the same turnaround we have now. Just that way, if a panel does get dropped by the fab, we have slack.

Cheers,

–David Carne

14

Hi, here be my bit:

Limit the orders, you know the square inch size of orders you can work with. disable checkout once the workload limit has been reached.

Ive been using batchpcb since it started (even before it started ;)) and know its possible.

@bigglez

Doubling the cost would mean we can go elsewhere for same price with 2 week lead times.

As for satifaction guarantee, sparkfun dont have control over these things, a fab house could do this, but not a 3rd party. IIRC the fab house do this to keep sparkfun custom, these odd shaped boards are probably more difficult to product than normal ones.

btw, $100 is a lot of money. Im a student and thats about 2 weeks food + 4 pack of beer. I (and my mates, of which i know 8 use batchpcb) would not a be able to afford this minimum order. Also my robots are 15sqr" perboard max, when considering that the motors for my robot cost $180 each (and Ive been saving for months to afford them) I simply cant afford a min order of that magnitude.

Meh, although i can get pcb’s made for next to nothing through uni if i need. But thats bending the system a bit. And i dont get emails from nathan enquiring about my funky shaped PCB.

Long live batchpcb!

15

busonerd:
I know what you mean by incentives - but in your case specifically, the order was filed with our fab [on time from our end!], and when we expected the tracking number to come in [about a 1 week window, which is how we’ve negotiated the pricing to be so cheap] we didn’t hear anything.

Dave,

Not sure who you are addressing above…

The problem, regardless of the deal (fees, delivery times) is that information does not flow back to the customer in a timely manner. I doubt SFE has much control over the fab (and any fab will try to push out ‘slow’ jobs to make a few $$ on a ‘rush job’). To get what BatchPCB needs out of the fab will take more effort from SFE, and that cost should be passed on to the customer.

Faced with “no BatchPCB” compared to a “more costly BatchPCB” I’d take the latter (while also shopping for a better deal in the open market).

busonerd:
At that point we started following up with the fab [1 day turnaround on all emails due to the time differential], and it turns out that somewhere in the fab the order got dropped [one we’d already paid for + sent to them]. At which point we said, “fix it then!” and then they started rush building the panel for us, and we expect it here any day.

Confirming my observation that these vendors operate by serving the customer that pushes. It the culture of that part of the world.

busonerd:
The reason the admin fee is “hidden” in the shipping, is because thats the easiest/only way to do it at the moment in terms of the shopping cart we use. For a while though, it should saying +10$ S/H to make it quite obvious its not part of the shipping.

The hidden fee is not a gating item in the disscussion. Unless profitable to SFE, BatchPCB will go away one day, anyway.

busonerd:
Raising the price would allow us to go to a more local fab + take shorter leadtimes, but for a lot of our customers, the response seems that it would be too expensive.

What happens to those customers when BatchPCB closes the doors?

Comments Welcome!

16

The reason the price is so low is precisely because we allow them to let them put higher priority stuff first [aka the 10 - 14 day window].

Gold Phoenix has had a pretty damn good track record in terms of shipments. The only regular slippages we have are ~ chinese new year [and we post notice of that]. Of all the panels we’ve done for batchpcb [well over 150] - we’ve had 2 that have been lost + we’ve had to have them rush make them.

We’re currently looking at ways of fixing up the information flow problem - I won’t have any more information on that for a few days, as it’ll take some coding.

Cheers,

–David Carne

17

pittuck:
Hi, here be my bit:

Limit the orders, you know the square inch size of orders you can work with. disable checkout once the workload limit has been reached.

Greetings pittuck,

That would be very silly… Do you suggest that the local gas station stop the pumps after, say, one hundred gallons have been sold?

We need to show SFE that BatchPCB can be a profitable venture. The sane alternative is that it goes dark, forever.

pittuck:
Doubling the cost would mean we can go elsewhere for same price with 2 week lead times.

Please indicate where you can get that service (PM me with the details if you like).

pittuck:
As for satifaction guarantee, sparkfun dont have control over these things, a fab house could do this, but not a 3rd party. IIRC the fab house do this to keep sparkfun custom, these odd shaped boards are probably more difficult to product than normal ones.

A fab house (any fab house) has to keep the chemicals warm. Any jobs they can fit in will be pure profit. What’s required here is a push from BatchPCB to keep their (our) projects moving forwards. If BatchPCB is marking the work flow by whether a shipment arrives on time or not, then there will alway be performance surprises. A pro-active BatchPCB would be asking ahead of time, not waiting in hopes of a delivery. Having delt with that part of the world professionally I know that is how their culture works.

pittuck:
btw, $100 is a lot of money. Im a student and thats about 2 weeks food + 4 pack of beer. I (and my mates, of which i know 8 use batchpcb) would not a be able to afford this minimum order.

A Ferrari is a lot of money - that’s why I don’t drive one. Your argument is weak. The only hope for us is to have SFE make a profit, without it we will always face the termination of BatchPCB.

I suggested $100 to get SFE a steady revenue flow. BatchPCB allows a one square inch PCB today, which could result in, say, $25 of gross revenue to service that customer.

Forcing the customer to come up with a $100 minimum removes that barrier. Your eight mates could share the $100 single project fee.

In other words the high cost and low margin task (concatenating several very small orders) has been pushed on to the customer, freeing up valuable BatchPCB labour to serve their larger more profitable customers. Additionally, BatchPCB has the assurance that each order generates enough to cover the cost of processing it.

I’m trying to remove barriers here - BatchPCB will go away if we don’t make some concessions as customers!

Comments Welcome!

18

We’ve come up with a solution - massive automation. Keep your eyes open in the next few days/weeks for more reporting info / etc. Most of the changes are behind the scenes, but it’ll allow us to cut human labour by a large amount. We hope with this we can keep prices the same, with only concessions in the worst case leadtime area, so we’re ok in case the fab drops another panel.

We don’t run BatchPCB for a profit. We do it as a service for all the hobbyists + designers on a budget. All of us here started off as electronics hobbyists [I was a customer long before I got hired here].

Cheers,

–David Carne

19

And as a hobbyistic approach to the service, and a service to hobbyists only; Setting a quota would not be the end of the world, and forcing me and my mates to group our order would incur even lower revenue (9x $10 = nearly a whole panneld PCB, which just one admin fee would not).

Anyhow, if the guys at sparkfun want some software for automation, give me a buzz with any ideas. I’ve still got the original batchpcb source code hanging around here (i use it to check me pcb’s before submitting it!).

20

pittuck:
and forcing me and my mates to group our order would incur even lower revenue (9x $10 = nearly a whole panneld PCB, which just one admin fee would not).

pittuck,

That is not what I said. If the threshold was $100, in my previous example, and you only needed $25 worth of PCB service, you could find others to meet the minimum. (That may be one or more or all eight of your mates, or just your projects).

In doing so you have unloaded BatchPCB from processing multiple orders for very thin margins. The single order that your co-operative submits is guaranteed to generate the minimum fees required to operate BatchPCB profitably.

Other companies impose minimum order fees.

Comments Welcome!