Margins & Pricing

[The Q]

Some of it definitely is because we can ..
Many internal to a cabin parts such as lights, switches, wiring etc are identical to those found in caravans , but if you buy them with marine labels, they'll cost several times as much.

Some of it is yachty snobbery, a well known manufacturer of car body filler (for sale in the UK and Europe) found it wasn't selling into the marine market, after some research they found it was because yachties don't trust anything that doesn't say marine on it. So they took their product put a marine label on it, charged more for it.. and sold quite well into the marine market.. At one time they admitted they had done that, they don't so much now as they are quite happy making increased profits.

 

[tane]

an anecdote that gives some indication about manufacturers margins:
the manufacturer in question (I was working for their distributor) sells directly to OEMs. (in the heyday of Bavaria this were 2 containers/week, "just on time")
Now a quite small OEM, who was our customer for other products, needed a product that had a longish leadtime at the manufacturer, but we had it in stock. Of course we had to give the OEM a really good discount, nearly purchasing price. They just laughed & said they want to just "borrow the product & give it back once they get it from the manufacturer.
Obviously the OEM price for a small OEM is WAY below distributor purchasing , so one can make wild-ass guesses about the prices for the big ones (70% below RRP?)
This BTW is the same for other "recreational" markets. Just see the retailprices of mtb-forks! For not so much more you get the whole bike.

 

[tane]

To show the fierceness of competition:
How does a manufacturer get a big OEM to change over to ones, say, rope-clutches?
They have to pay for a new mold for the deck, if the clutches don't fit the existing deck. & of course the new clutches have to be cheaper!

 

[Ease the sheet.]

If you ad ethernet, you need a new back to accomodate the ethernet jack. If you then add N2K or seatalk, you need to mod the case. Etc. Its a nice thought. But harder than you think unless the product design is completely frozen.

Or you can just design a case with a plug in section that allows upgrades.
Making a whole new case because you want to change a plug is silly

 

[Monkey]

Or you can just design a case with a plug in section that allows upgrades.
Making a whole new case because you want to change a plug is silly

Now you’re paying for two injection molds instead of one, and you still need a new mold for each plug change. There’s no free ride in mass production.

 

[Old Pirate]

I spent the last 10 years designing consumer electronics devices, essentially specialized tablets with WiFi, LTE, and a bunch of other stuff added in. I spent a lot of time in China at the contract manufacturer (you've probably heard their name in the news and used a product they built) working with the engineering and manufacturing teams to ramp up manufacturing on each product. At 500K units/year we were a very small project; the popular cellphones, laptops, etc. start at 500K/month. A new high end phone can ship 10M in the first few months, and then there's the lifecycle treadmill where it gets a "refresh" at about 9 months, and then replaced by a new generation at about 18 months. We shipped essentially the same product for about 5 years until the key chips were end of life and we had to design a new generation.

The points above about manufacturing volume and product lifetime are right on. Toward the end of the product life cycle we were constantly chasing supply for key parts because after a while the manufacturers move on to the next generation of technology and don't want to bother building low volumes of the old stuff. Not just micropocessors, but everything: memory, LCDs, cameras, even capacitors that cost 0.01 cents each.

Designing for upgrades sounds logical and easy but in practice it's not. Building in the necessary capability to allow the customer to add an option down the road costs money. Connectors on the PC board, capacity in the power supply, space in the enclosure, punch-outs in the case for the connectors, etc. all mean extra cost for the base unit. In a low volume device with careful design sometimes you can make it work, but in the end you are putting an added cost burden on every unit to support the small percentage of customers who might add a feature in the future. Of course PCs have been doing this since the beginning over 40 years ago where has been the concept of the motherboard with add-in cards, and even then you have to choose your power supply, case, cooling, etc. to support your future desires. But that's a rare exception.

You might hang on to your cellphone for 2-3 years, or if you're a curmudgeon like me 3 generations back is just fine and you'll have it for 5 years or more. But I expect the 5 year old B&G system on the boat to keep going and have parts available for another 10+ years. That's tough for the selling company to keep supporting with parts, software upgrades, backwards compatibility in new devices, etc, while funding engineering for new product development and paying for the rest of the company overhead.

 

[Ease the sheet.]

Now you’re paying for two injection molds instead of one, and you still need a new mold for each plug change. There’s no free ride in mass production.

Technically your only paying for a second mold...

Though a mold with removable/replaceable segments isn't difficult.

3d printing could even play a role.


Don't get me wrong, I understand the position manufacturers are in. My opinion is that the "marine" market is used to paying over and above other markets.

And nobody wants to kill that golden goose.

 

[DDW]

I am quite familiar with both big and small manufacturing of electronics. There is a strategy for some marine electronics that isn't being pursued: modularity, allowing mid cycle upgrades and piggybacking on consumer volumes. I kinda get why the big names don't do it - their money is in your pocket, and their job is to free it up.

For example, if you could buy a 1000 nit waterproof touchscreen, you could upgrade the electronics driving it several times before the screen would be obsolete enough to want to replace it. The horsepower in the computer section on the other hand, even in high powered MFDs is a fraction of a midrange cell phone. The cell phone probably costs less at retail than the manufacturing cost of the marine computer board, because of the 6 or 7 orders of magnitude difference in volume.* And the cell phone speed doubles every year or so. Conceptually, you buy electronics intended for high volume consumer use and retarget it at marine. Your hardware R&D costs just went to zero and your per unit costs drop by half. The OS including all the graphic slinging, networking, UI is already done, maintained, and upgraded for nothing so your software R&D is cut in half or less. Of course there are issues with this approach, but no one has really tried it. We are starting to see a breakdown with radar for example becoming generically displayable on open sourced software. And MFDs running Android as an OS.

* Over on a machining forum a couple of years ago when the Apple watch was introduced along with some ad copy claiming that Apple had developed a new aluminum alloy for it, a lot of derision was heaped on them for that sort of outlandish claim. New aluminum alloys are developed for Boeing, not Apple, it was said. Until someone pointed out that during the 4th quarter, Apple was shipping the equivalent weight of one B-767 in aluminum every 20 hours in cell phone cases. Boeing's consumption of aluminum was a tiny wart on Apple's ass.

 

[Ease the sheet.]

I am quite familiar with both big and small manufacturing of electronics. There is a strategy for some marine electronics that isn't being pursued: modularity, allowing mid cycle upgrades and piggybacking on consumer volumes. I kinda get why the big names don't do it - their money is in your pocket, and their job is to free it up.

For example, if you could buy a 1000 nit waterproof touchscreen, you could upgrade the electronics driving it several times before the screen would be obsolete enough to want to replace it. The horsepower in the computer section on the other hand, even in high powered MFDs is a fraction of a midrange cell phone. The cell phone probably costs less at retail than the manufacturing cost of the marine computer board, because of the 6 or 7 orders of magnitude difference in volume.* And the cell phone speed doubles every year or so. Conceptually, you buy electronics intended for high volume consumer use and retarget it at marine. Your hardware R&D costs just went to zero and your per unit costs drop by half. The OS including all the graphic slinging, networking, UI is already done, maintained, and upgraded for nothing so your software R&D is cut in half or less. Of course there are issues with this approach, but no one has really tried it. We are starting to see a breakdown with radar for example becoming generically displayable on open sourced software. And MFDs running Android as an OS.

* Over on a machining forum a couple of years ago when the Apple watch was introduced along with some ad copy claiming that Apple had developed a new aluminum alloy for it, a lot of derision was heaped on them for that sort of outlandish claim. New aluminum alloys are developed for Boeing, not Apple, it was said. Until someone pointed out that during the 4th quarter, Apple was shipping the equivalent weight of one B-767 in aluminum every 20 hours in cell phone cases. Boeing's consumption of aluminum was a tiny wart on Apple's ass.

And don't get me started on proprietary software and the "software as a service" model...

 

[Old Pirate]

...Conceptually, you buy electronics intended for high volume consumer use and retarget it at marine. Your hardware R&D costs just went to zero and your per unit costs drop by half. The OS including all the graphic slinging, networking, UI is already done, maintained, and upgraded for nothing so your software R&D is cut in half or less. Of course there are issues with this approach, but no one has really tried it. We are starting to see a breakdown with radar for example becoming generically displayable on open sourced software. And MFDs running Android as an OS.

That's exactly the strategy my company followed for the last 10 years. It works, but only for a year or three. The problem is the consumer market for phones and tablets moves so fast that this year's high volume part is tailing off next year and going EOL the year after. Same for the OS - Android revs every year and even if you are on the chipset platform with the long life cycle support, the manufacturers will only go a few versions before they stop. For example if you started a design in 2020 with Android 10, Qualcomm has a few chipsets they guarantee availability for 5 years and support to Android 12. But by 2025 we'll be at Android 16 or 17 and the chipset from 2020 that was a 12nm process has been replaced by a 5nm process and TSMC wants to shut down the 12nm line. The 8MP camera will be EOL because the high volume point is now 20MP and the interface is different, memory has gone from LPDDR4 to LPDDR5 to LPDDR6 or whatever's next.

There's definitely savings in HW R&D but the reference design only gets you to about 90% and then there's the second 90% of the project to do the necessary customization of the PCBs, enclosure, antennas, etc. Using that leverage, we had a hardware team of 5 keeping a SW team of over 100 very busy. SW has a similar problem, you get the base OS with networking and graphics and then you have to put your own app on top of that. Unless you have a team big enough to keep up with the Android revs while maintaining the product on the current rev you get stuck on whatever rev you started with, when your chipset is EOL and you're forced to move to a new platform there's a huge effort to get to the new rev.

All that said, the challenge is less in the engineering than in finding components and manufacturers where you can get a long enough guaranteed availability. Most of the low volume niche markets are resistant to change and slow to take up new tech so you need to have 5 year minimum life cycles instead of 12-18 months and that's tough. It's funny, I've had a customer tell me he trades in his phone for the latest flagship phone every year, and then in the next breath say "I loved the old product, how come you're making that EOL and forcing me to buy the new one? And it's too expensive!"

What a lot of the niche manufacturers are doing is running Linux on silicon that isn't aimed at the consumer market - you get the advantages of a stable OS with networking and graphics running on silicon that has longer life cycles.

 

[DDW]

Those are the 'issues' I mentioned. Yes you have to do the app. Yes you have to do a wrapper for the hardware. I used cell phones as an example, however they are not the best start for a chartplotter. It isn't a particularly high powered task, as an example OpenCpn runs pretty well on a Raspberry Pi, around $50 - $70, 1ea at retail. Runs Linux with very good legacy support. It can plug directly into a HDMI touchscreen monitor with nothing extra. Canbus (NMEA2000) interface is off the shelf (or just WiFi it). Already has WiFi, Bluetooth, and Enet.

If I were every to build another custom boat, there would be very little proprietary marine electronics onboard. Radar scanner and VHF w/AIS is about it. I probably would not go the dedicated Raspberry or other route, just a brick PC (or Mac). My last custom was launched in 2008/2009, chartplotters have been replaced twice at considerable expense. The original ship's computer (Mac Mini) still runs fine and still runs the charting software. Even if replacing that twice, the cost to do so would be a small fraction of what the chartplotters have cost. The disadvantage of that is it is a kit, I have to say that modern MFDs and instruments do integrate very well without needing to know much. That matches well with the market, which is basically redneck fishermen with few computer skills. Sailors are in a way fortunate to piggyback on that much larger market.