Opinions on a new business gratefully received...

[angust_uk]

Hi all

I hope this is an appropriate place for this post - if not I'm happy to remove it (although I'm not soliciting for business - just opinions!).

I'm after some feedback on a very early-stage business idea I've been working on and would very much appreciate any feedback, comment or criticism from the hive mind. To give some background, I'm an EE with 10+ years' experience and got a bit tired of working for other people. Instead I'm mulling over building a web site that would allow EE's and non-EE's alike to turn their schematic or prototype hardware into a professionally designed PCB, as well as taking on (well, subbing out) the fabrication and assembly.

I'm aiming to target 2 broad categories of people (or companies) - those who already have experience in layout and manufacture but who don't have the time or capacity to do it themselves, and also those who aren't electronic engineers themselves but have the need for something a little more polished than a breadboarded design.

I imagine most of you fall into the first category and I'm hoping that at some stage of your working life you may have found an end-to-end, fast turnaround service like this useful. I'm aiming to offer a very quick service (with an appropriate premium added if this is the case) for those projects which need hardware built yesterday but perhaps the company in question doesn't have the capacity to do it themselves.

Obviously there are many existing solutions to designing and building custom boards (I'm sure some of you do this for a living!) so my aim is to create a very straightforward interface to allow upload of data (from a sketched block diagram to fully fledged schematic) and fast turnaround quote (aim is within half a day during working hours). We (initially just me, but hopefully not for long) would then take on everything from layout and DFM to procurement and assembly and courier your completed boards to you.

This post is already far too long (if you've got this far, well done!) so I'll leave it there, though feel free to ask questions.

I'd very much appreciate any comments you might have on the viability of this - as I said, the idea is in its infancy and very much subject to change!


Many thanks in advance

Angus

P.S. If your feedback is 'that's a stupid idea' absolutely fine, but I'd far prefer 'that's a stupid idea because...'

[sokoloff]

In doing some mechanical design for 3D printing for others and in discussing a few maker-type ideas, I've been surprised with how little people expect to pay. In the custom design and 3D printing market, the reference price is "cheap injection molded crap". In the maker space (a place where I'd expect a little more understanding of what's going on), there was still a reference price of "Raspberry Pi is a whole computer".

In that regard, I'd definitely target businesses and not individuals as your sweet spot. There's nothing to prevent the occasional individual from finding you and buying your services, but if you target individual, early stage maker/dreamers as your sweet spot, you're going to find a lot of people who don't have money to spend. Targeting businesses may have a longer sales cycle, but is likely to result in more reasonable business plans, a better understanding of the costs of people and benefits of outsourcing, and more likely repeat business over the years.

Good luck! 

[angust_uk]

Hi sokoloff

Thanks for the feedback, and for the record I completely agree! Having offered custom design work in the past (several years ago now) we scared off many individuals by quoting a reasonable rate (by reasonable I just mean enough to pay the bills and very little else!). I'm definitely targeting small companies who may not have in-house staff or larger organisations who may do already but are likely pretty slow due to internal processes etc.

Cheers

[MK14]

There's (or can easily be) huge costs in running and starting up a business in the UK. Some of these costs are not at all obvious.
So, I'm not sure if there would be anything like the amount/volume of business, to recoup those costs, let alone make a profit.

So, I'd suggest you be very careful, before giving up on a (presumably) successful career.

The old saying goes, "The grass is always greener on the other side".

[cdev]

When estimating potential market share and customers, always take changes in the market into account. Also, new technologies. A new technology can come along and wipe an entire business out globally in a few years.

Also, keep in mind that services, ("everything you cannot drop on your foot") maybe 75% of the economy, and most of the better jobs - are being liberalized by the WTO and other global entities (see bilaterals.org for a list of bITS, for example) "opening" many fields to competition, but in reality making it prohibitively less profitable forfirms to operate in the more expensive countries unless their product is totally unique). If your firm is large and you are in  a low-regulation low-wage country, and your costs are much lower than in other countries, that puts you at a huge advantage if you export the service. Indentured-servitude-like worker arrangements may also make it possible to use your workers from home abroad. At least that is the hype, hype which appeals to many resellers of various services . Rules have been changed or are being changed (by the WTO and in other trade bodies, respectively) to eliminate the red tape that has historically stood in the way in individual countries, as well as public services which are framed as state monopolies preventing businesses from the profits they are entitled to. (Countries that maintain professional public services have a high level of risk under these agreements which have a plethora of traps for the unwary. Public services of all kinds are being demonized as 'protectionism')

These changes are binding on countries that sign on to these groups, and may be demanded as a condition of entry. (In the case of the WTO, easily 90% of the countries in the world.) 

They could potentially turn the work world upside down, also making good quality professional work much harder to find and keep, but larger businesses more profitable, in theory, by lowering worker-related costs. In theory, the chaos would work to the benefit of firms that resell those services across international borders in various ways. Or so goes the hype. But, who will buy the services?

These changes will almost certainly work to the advantage of huge multinational firms that locate where wages are cheap, and put a lot of smaller firms that cant join the race to the bottom, as well as unions and likely many individuals, out of business.

[AndyC_772]

I work as a consulting engineer, and given past experience, I'd politely decline (or, perhaps, renegotiate) some of the jobs you're proposing to take on.

The problem is that, when you deliver a prototype (or schematics, or layouts, or whatever), you're implicitly responsible for making sure it works. If it's completely your own design then that's fair enough, but the difficulties start when the customer's supplied materials have errors in them, or just don't work the way they expect.

This means you end up having to carry out a full design review of everything you're supplied. It could be that you're given a perfectly good prototype which really does "just" need laying out on a custom PCB - but, equally, a design you're given to work from could be a disaster in terms of reliability, or EMC, or production yield, or cost, or technical performance, or whatever. This becomes your problem as soon as you produce a derivative product that inherits the same problems, and you'll end up taking the blame for not solving them.

Bear in mind that if your customer understood those problems in the first place, you wouldn't be having to deal with them. You may find yourself having to explain to someone that their own work is bad, and that's all but impossible to do diplomatically.

With that in mind, I have a policy now that I work from specifications, and I do the design work myself. This means I'm only responsible for my own choices, and inevitably it's quicker and more reliable in the long term if I'm developing and maintaining something that's done exactly how I'd have done it, and which has no unnoticed 'gotchas'.

All my customers are businesses who have manufacturing capability, but not electronics expertise. They hire me on a regular basis whenever they need something electronic designed, debugged or modified.

I like it this way; we all work well together, and everyone's time is used efficiently. If you deal with too many people each of whom only wants a one-off, you'll find yourself spending a lot of time and effort dispensing free advice.

[station240]

Instead I'm mulling over building a web site that would allow EE's and non-EE's alike to turn their schematic or prototype hardware into a professionally designed PCB, as well as taking on (well, subbing out) the fabrication and assembly.

...

Obviously there are many existing solutions to designing and building custom boards (I'm sure some of you do this for a living!) so my aim is to create a very straightforward interface to allow upload of data (from a sketched block diagram to fully fledged schematic) and fast turnaround quote (aim is within half a day during working hours).

There is no way you can have half a day (4 hours or ?) turn around on converting some random schematic or breadboard/rats nest prototype into a PCB.
People will send you anything, files from obscure schematic design programs, stripboard with point to point components and jumper wires everywhere, scraps of paper poorly scanned and originally drawn in pencil. Someone is bound to send you a design with a 200 odd pin QFP 32 pin micro or worse something with BGA.
Even ignoring those issues, it's not practical to do every PCB layout in a few hours, making the device footprint when it's in no one's library, and checking the datasheet recommended PCB layout could take that long.

Being the middleman between the customer and PCB factory is also going the prove a pain at times. You have to consider which PCB house can handle say 5 thou track spacing, or chase up the issues that arise when plain bad luck means they have problems with your design/files*.

* I went though a lot of issue with my last PCB run, all of it seems to be at their end, but I just don't know as they fixed it without explaining the issues.

[awallin]

2 eurocents:

1. we asked a few companies to assemble open-hardware with 0.5mm (iirc) BGA packaged fpgas and ram-chips - so far most replied they won't do it because they can't verify the bga-soldering (x-ray machine required?) - since more and more things are BGA maybe that is becoming standard even for low-ish end boards?

2. more and more boards contain software and firmware. you'd have to master that side also to some extent (vhdl, jtag, uControllers, etc) in order to deliver complete products with more added value to a customer. As mentioned before in this thread, think about who is responsible when the software doesn't want to talk to the hardware: the schematics-designer, the pcb-assembler, the firmware-programmer, or someone else? How do you divide responsibility between the customer and the assembly-house?

[angust_uk]

First off, many thanks for all the very useful and carefully considered replies - you've given me a lot to think about.

There is no way you can have half a day (4 hours or ?) turn around on converting some random schematic or breadboard/rats nest prototype into a PCB.

Maybe I wasn't very clear - I'm certainly not promising to turn around an actual layout in half a day (unless it's very simple, of course) - I was referring to the time to get a quote back to a potential customer.

1. we asked a few companies to assemble open-hardware with 0.5mm (iirc) BGA packaged fpgas and ram-chips...

2. more and more boards contain software and firmware.

Very good points - regarding the firmware this will likely be something I'd look to offer in the future, but certainly not initially

The problem is that, when you deliver a prototype (or schematics, or layouts, or whatever), you're implicitly responsible for making sure it works.

Yes, this is definitely one of the (many) bigger risks - in the first iteration I would be clear that, given a supplied schematic, the board produced would guarantee to match that schematic, working or not (obviously worded in a slightly more diplomatic way). Longer term I would definitely have to carefully think about taking on this risk and what I could potentially 'guarantee'.

These changes will almost certainly work to the advantage of huge multinational firms that locate where wages are cheap, and put a lot of smaller firms that cant join the race to the bottom, as well as unions and likely many individuals, out of business.

One option I've been considering is taking advantage of the global 'gig economy' to allow engineers worldwide to bid on some or all of this work (in a similar way to Elance or freelancer.com), but with an added review process to ensure (as much as possible) high quality output.


Thanks again all!

Angus

[cdev]

Be careful, Angus and keep in mind that you're likely not the only one planning to do this.

We're starting to see another bubble develop, its developing because of huge numbers of people who want high yield investments, not because there are a great many really great, unique ideas.

Its being done for a reason, because the truth is actually pretty scary. We don't know what will be happening a few years out into the future. Its increasingly unpredictable. Things are changing so rapidly.

 the Digital Economy is being hyped as a solution to almost everything, a virtual Eldorado of economic dream fulfillment, where corporations from the poorest and least equal countries especially, because of their consistently low regulatory burden, and low labor costs make large profits reselling the services of their large numbers of graduates.

However, its a pie in the sky situation to prevent global realization of an uncomfortable truth.

Big shifts are occurring in how money is made and by who. Humanity is facing a big challenge in the form of automation of all kinds.This will concentrate wealth so intensely at the ever shrinking top and quite likely lead to a race to the bottom on wages for everybody else. Putting even more pressure on businesses to economize.

That said, I wish you the best of luck. We live in interesting times, and the potential for people to make a difference is larger than it might seem, especially if they think originally.

[SiliconWizard]

Nice idea although business is probably going to be very tough.

Just forget about ultra-fast turnaround IMO. That would be shooting yourself in the foot big time. I know it sounds like a nice selling point, but unless you're *over*-staffed, it's never going to work out and may end up killing your business. Do not promise that upfront. Just let your customers ask for it if they require it, and then only accept once you have gathered enough info on the project and know for sure you're going to make it.

Another point that comes to mind is that you're going to have to invest a lot of time and money in CAD tools. Potential customers WILL use all kinds of CAD tools and WILL often require that you use the same. This point itself may cause a big issue for a starting business.

[mikeselectricstuff]

regarding the firmware this will likely be something I'd look to offer in the future, but certainly not initially

Everything has firmware these days. You need to at the very least be able to write enough  to verify that all the hardware works.
Most "non electronics" customers will also probably have no software resources and expect you to do all of the firmware as well.
If you can't do this then a big part of your potential market is gone.

I can't immediately think of a single job I've ever done that hasn't involved me writing some firmware, and/or also support software on a PC for testing/exercising/calibrating/setup/logging etc.

If you do have a client that can do software, there's a good chance that at some point there could be a dispute over whether a problem is their software or your hardware. More so if their "software person" is inexperienced in embedded stuff. I have had a customer who called them selves a programmer ask me to explain how binary numbers work. Seriously.
   

[sokoloff]

+1 on Silicon Wizard's first two paragraphs!

On the third one, I think many of your customers won't have any opinion on CAD tools and the ones that do, you'll have to discuss the costs to them for you to comply individually. It's maybe going to be an issue here and there, but I wouldn't let that stop me.

[AndyC_772]

Yes, this is definitely one of the (many) bigger risks - in the first iteration I would be clear that, given a supplied schematic, the board produced would guarantee to match that schematic, working or not (obviously worded in a slightly more diplomatic way). Longer term I would definitely have to carefully think about taking on this risk and what I could potentially 'guarantee'.

Make sure you're clear about what the service is that you're providing.

If someone comes to you with a schematic, and you produce a PCB out of it, then you're a PCB bureau. There's nothing wrong with that, good ones are in strong demand and I know of one which is hiring right now.

The next stage in terms of value you can add, would be to transcribe and review schematics as well. This puts you into the realm of being a full turnkey design service, and you really can't do this without offering guarantees that the things you do will work. Bear in mind always that people will hire you for your expertise, and whatever you do (or don't) promise, you'll get repeat business if you deliver working products on time, and not otherwise.

Don't worry about any technical issues, like having to place BGAs or use unfamiliar CAD tools. You're already an experienced engineer and you'll be able to anticipate these kinds of things and deal with them sensibly. They're not the problems that will keep you awake at night.

The real problems are the ones you don't encounter as an employed engineer that you will run into when you're running a business in a customer facing capacity.

For example, you'll run into people who have different ideas to you about who owns what in terms of intellectual property. If I have one piece of advice for you, it's to hire a good IP lawyer for a few hours before you start, get them to review your T&C's of business, and make sure you confidently understand how the law works in terms of copyright, IP ownership, licensing and liability - at least as far as they apply to the service you're providing.

People may try to get you to sign your rights away. They may act dishonestly or incompetently, and you may struggle to tell the difference. They may assume that you're guaranteeing things that you're not, and it may not even occur to them to tell you, even if you ask. Some customers will work with you in a mutually beneficial way, others may genuinely think they have the right to own you and everything you do for as long as you have anything to do with them.

[TheNewLab]

How about a better version of FaceBook that does not track you, of capture, track and analyze all of your information?

HQ it outside the USA? like well, the EU with there privacy as a human right?

Watch Frontline two part episode on "The Facebook dilemma