[mtlynch]

Thanks!

>Wondering what your thoughts are of these extremely low cost kvm’s from Sipeed (NanoKVM)?

>Do you think that allows you to expand your market since the hardware is cheaper as you maintain great user experience? Or does that force you to go upmarket as hobbyists need only the minimal feature set?

To be clear, I'm totally gone from the company at this point, so I'm not thinking about strategy for TinyPilot at all anymore.

But I will say that every year, I'd see a new KVM over IP pop up that claimed that they were going to undercut TinyPilot by 60%. And then they fizzle out, and I never hear from them again.

My suspicion is that people see TinyPilot and say, "Wow, that looks like $100 worth of hardware being sold for $400!I could do what they're doing and sell for $200 and eat their lunch and still make $100 on every unit!" But then as you get into it, there are all these more subtle costs like compliance testing, tariffs, customer returns, insurance, etc.

And that's before you even get to customer support. For a KVM over IP, you can't just give customers a "have you tried turning it off and on again?" support response because the issues are more technical and deal with things like NATs and proxies. So if you're making $20 per sale on a low-cost device, and then the customer has one conversation with a support engineer, you lost your profit and probably would have been better off not selling to them at all.

So, I think there's room to reduce prices as hardware prices come down, but I'll be surprised if other vendors can slash prices to below $100 and still run profitably in the long-term.

[restalis]

Can't those "more subtle costs" be pulled out of the product cost itself and offered or at least expressed separately? Say customer support, if it'd be offered as a paid service instead of "free" as in baked-in-product-cost, thus giving the customer a choice to either go for that easy-pick fruit of available paid support, or the alternative of investing more of their own effort into figuring out and solving the encountered problems by using the docs available to them, or maybe giving up on the problematic use-cases in the first place? Or tariffs - is it bad to let the customers know that there are cost differences in the product offered to them due to their request coming from different tariff-impacted markets?

[mtlynch]

>Say customer support, if it'd be offered as a paid service instead of "free" as in baked-in-product-cost, thus giving the customer a choice to either go for that easy-pick fruit of available paid support, or the alternative of investing more of their own effort into figuring out and solving the encountered problems by using the docs available to them, or maybe giving up on the problematic use-cases in the first place?

My sense was that the customers would have been pretty unhappy about that setup. Like, a common question was, "I plugged in my device, but I don't see it on the network. What do I do?" They'd be pretty annoyed if I had just said, "I'll help you if you pay me $99 for customer support."

I think in practice, if I tried to pull that, the customers would just say, "Okay, then I'm sending it back for a refund," or, "Okay, then I'm calling my credit card company for a chargeback," both of which are costly for the vendor.

Plus, I personally don't like the dynamic of customer-paid support. If customers have to pay for support, it incentivizes the vendor to make the product difficult to use so that customers need support.

I always want my products to work so well that customers don't have to contact support. So, if a customer has an issue, I want to feel the cost of that as the vendor so I'm pressured to fix it.

>Or tariffs - is it bad to let the customers know that there are cost differences in the product offered to them due to their request coming from different tariff-impacted markets?

For tariffs, I actually meant tariffs that I pay as the manufacturer. For most of the time I was running the company, we were sourcing most components from overseas and doing final assembly in-house, so we paid tariffs on components as they came in. That's a huge headache, especially when we have distributors in other countries, and we're paying tariffs multiple times for multiple border crossings. And tariffs seemed to change per shipment by 2-3x for reasons that were never clear to me.

Tariffs that the end user pays are a whole other headache. We didn't have a way of predicting what tariffs the customer would pay, so customers would sometimes be upset that we didn't warn them, especially in countries when they receive the tariff bill after they've already received the item. At one point, we tried to use DDP, which is supposed to let us pay the tariff on the customer's behalf, but we tried it a few times and the customer was charged a tariff anyway, so it seemed like it was just a huge expense that did nothing.