Hacker News Reader

Comment by iamjdg | Hacker News Reader

So they are basically renting their pumps? When the pumps fail, they must get them back and resell them to another customer. Else the economics would not make sense. Unless they send less to build a shorter life pump, but my impression was no, they program it to fail after a certain number of operating hours.

View on HN

The economics make perfect sense. Imagine you have a pump that lasts 1,000 hours, and it costs $10 to build. You have a reputation for making great pumps.

To recoup your R&D costs and make a profit, you decide that you're going to sell these pumps for $100. Those in the market for a 1,000 hour pump think that's a fair price, but those who only need a 100 hour pump think that that's outrageously expensive.

If you bring the price down to make the 100 hour pump people happy, you're cutting into your margins for the 1,000 hour people. So you segment your market: $100 for a 1,000 hour pump, $40 for a 100 hour pump. Your 1,000 customers continue buying the pump at the price they've always been paying, and some of the 100 hour people start buying at $40. That incremental revenue is all gravy; you didn't have to sink R&D money into a second product (other than building the DRM)

thanks. yes you are right. it is market segmentation. you move more units for a lower weighted average margin. just seems silly your actually modify the product to make it fail early just because someone paid less for it. i get it for software, you leave features off or put them on and charge less or more. but for an actual physical product, you are playing with operating hours as a feature, just seems new to me. if i was to do it, i would put some money into picking cheaper parts for the lower lifetime pump. then i could perhaps build it for $7 and get more margin for these sales. but i guess it depends on how much money you put into integrating cheaper parts. but even making it fail early by programming cost some money.

I agree, it's totally weird. I've been thinking about it ever since I saw the article, and thinking about other pieces of equipment that I have that operate on a similar model.

I've got an oscilloscope that I bought as a 50MHz model, knowing that an upgrade to 100MHz could be purchased after the fact. When I bought it, I didn't need the extra bandwidth and it was a decent price. Recently, I started to need the extra bandwidth so I inquired about the upgrade cost. $300. To do the upgrade, they give you a file that you stick on a USB key.

Part of me is grumpy about the fact that the equipment I bought was actually capable of 100MHz out of the box and was just software locked. And part of me is happy that I got to hold on to my extra $300 until I had a compelling reason (read: a client was paying) to do the upgrade.

$X was a decent price for a 50MHz scope, and $X + $300 is a decent price for a 100MHz scope.

i still think they would make more money renting them to the 100 hour customer, bringing them back, then selling them to the 1000 hour customer at a reduced priced because it is used and only a 900 hour pump.

Really depends on the manufacturing cost of the pump vs. the cost of shipping it. One of the really interesting things for me with the rental model is that if they kept the timer circuitry in it, they'd be able to collect a LOT of wear data. You can do in-house testing, but getting access to that level of field data is hard.

yeah good catch, i didn't think about it, that data would be gold.

Reply on news.ycombinator.com