In this workshop‑style discussion, Sean Hsu shares practical advice for early‑stage hardware startups on building a minimum viable product (MVP) and moving toward market traction. He contrasts quick‑launch website tools (Wix can be set up in 1–2 hours, WordPress in about two weeks) and urges founders to avoid spending months on a site before validating demand. Hsu recommends filing a provisional patent when needed, outlining cost ranges ($200‑$500 for DIY filing, up to $1,500 for attorney assistance) and tailoring protection strategy to target markets (full patents for US/EU/JP, self‑drafted filings for China). He mentions using KeyShot for 3‑D rendering of product designs. For go‑to‑market, Hsu cautions that crowdfunding (e.g., Kickstarter) can generate early sales but does not guarantee sustainable growth; instead, founders should allocate $3‑$5 k to paid ads (Facebook, Instagram, Google) to identify the broader “ocean” of potential customers beyond the initial “fish” pool. He then dives into unit economics, describing a concrete hardware example: a $500 product with a target bill‑of‑materials (BOM) of $125 (25 % of price), 20 % overhead, and a maximum 60 % marketing/sales cost. He calculates customer acquisition cost (CAC) as cost‑per‑click divided by conversion rate, aiming for CAC to be roughly equal to the product’s gross margin (e.g., $100‑$120 CAC on a $369 price point yields 30‑40 % margin). Hsu argues that early‑stage hardware should be priced premium (citing Nest thermostats, iPhone, Tesla Roadster) to secure higher margins that can absorb production errors and allow flexibility. Throughout, he uses analogies (dating multiple “girls” to set hard deadlines) to stress the importance of rapid iteration and cash‑flow milestones within three years.
Chapters
So last thing for the MVP. For the website, how do you guys build websites now, by the way? What kind of tool do you use to build websites?
WordPress.
WordPress. How long do it take to launch a website from zero to launch?
2 weeks. 2 weeks.
Okay, let's try— I mean, we definitely should have to go to Warp Drive for production. I think that's a more legitimate solution. But for MVP, I typically use— so this website I use Wix. It's like also 1 to 2 hours work. I'm trying to get like— so in this phase, again, in my chart it was like a big mess up and then it all become clean. What I want to avoid is I don't want any entrepreneur really spending 2 months to build a website that they don't even know this is going to work at the very bottom point yet. My approach is typically I use a lot of tools to spend 1 or 2 days to build an MVP and just spread out all my MVPs and see who picks up what. Sometimes that's not enough time. I use a very bad analogy that it's like picking up girls. You don't necessarily just go for this girl, this is my true love, I'm going to go for it forever. Typically if you really want to get married in a certain age, let's say assuming if I was single and then I only have 3 years and I need to date someone and get married, just assume that's a hard requirement.
Right.
The same thing for entrepreneur. You have to set up a very hard deadline for you. After 3 years, you have to be earning cash, otherwise you will just go back to work, right? So you typically just try to flirt with 20 girls.
That's stressful.
OK, I think you got the point. So that's my point. So try Wix. Wix is pretty easy to use. Yeah, go ahead. For patented, I go for provisional patent first if I feel like.
Like this delicate machine, I can see myself using it. Yeah. But like you already made the website, it's beautiful and I'm sure you've not like launched it, but how do you protect that design to be yours other than website?
Well, if my market is US and Europe and Japan and countries that they respect patent, then I'll get a patent. But if my target— if I'm selling to China, I'll just write the patent by myself. Spend an hour just describe all the patent. So it's basically like— You can find a template of like a provisional patent application and then spend $200 to $500 to file by yourself. Or if you really need a better lawyer to write it down, I know some lawyers, they can only cost $1,500. So once you have the draft, they will clean up the draft for you and then they'll just protect it for you. What does he want?
Can't you find your own professional factory for this design?
For this design, actually, this is not the final design, so I don't care. This is just something that looks fancy. So my hypothesis is that if someone knows— if there's a market of this kind of machine, then there's a hope. That hope is enough for me to spend more than $5,000 and more more than a month to develop the actual product, get some technology done, and also get the patent. Go ahead.
What tool did you say you used to make the lighting?
KeyShot. KeyShot, okay, thank you. So the last point about this is that I— Sometimes I believe— so I think at this point there are 2 ways of success in a new startup. You can either get very successful by Kickstarter or any crowdfunding, but crowdfunding don't actually secure your future growth. This is like fishing, right? Kickstarter, you know this pool of fish is always there, there are new adopters. actually adapting anything new. Anything new or innovative, they will just buy it for cool or show off. But what you want to see is the whole ocean. There are so many different fish in this ocean. You want to try like who is actually going to buy your product eventually. So I know a lot of good friends, they get through all kind of incubators. I'm actually mentoring at HAX and Highway One and all kind of incubators here. So a lot of startups, they actually fundraise, Kickstarter success, fundraise more, and then die. The reason is because they could not find the actual fish in the big ocean. So I definitely recommend go for crowdfunded, but also at the same time, you really need to find what kind of fish in this ocean you're looking for. And by using that, It's not that hard, but it's more like spending— this is more budget. It's like $3,000 to $5,000 to spend in Facebook ad, Instagram ad, and Google AdWords. And just get more understanding about who is actually going into your website, who is actually going to click the buy box, and who is actually going to—
Buy.
So what's your—
from here to validation, do you have metrics that you seek for?
So my metric— very good point. My metric is if I'm selling this machine— I'm just using machine 2, for example, but I'm not actually working that much on this now. If this machine is $500, my BOM has to be A fourth of that, so that's like $125. My operations is typically like, I'll call it 20% for overhead. So if my marketing is, or sales is more than 60%, that's going to be very hard because that is what channel will cost for me. Like if I talk to Best Buy, they'll definitely buy this—
$50.
Like $250 or something. So if my cost is already like $200, I'm going to earn 50%. I mean, it's still a pretty good margin as like a business. It's like 25% margin, but it's not like a huge margin for a startup. So my metric is typically if the bomb is 25%, I'm hoping customer acquisition cost is 25%. So not the lifetime value, it's the actual customer acquisition cost per one-time purchase. So that is the equation of the click per— sorry, cost per click cost divided by your conversion rate. And that's the acquisition cost. For this one, for example, it's around $100 to $120. So it's a little bit more pressure for me. So actually, My experiment on this guy was like I actually ended up finding customers are willing to pay $369. The bump cost is $120. The acquisition is $100. So I'm only earning, if I sell directly online, I'm only earning like 30% to 40% of margin, which is putting myself in a very tight position. I don't want— I want to be in a more— What's the bump cost? higher margin business, that kind of mentality. So that's one of the reasons I didn't move forward with this. What kind of margin are you looking for? I think for SaaS, I think more than 50% will be more safe. Because for example, if you look at how Tesla success, we start with a very expensive sports car, Roadsters, that is like $150,000 to $200,000. So there's a lot of margin for us to— or there's a lot of money there for us to tolerate with the error in the startup. The errors such as like you find out the whole batch of the T0 or T1 injection molded are like so totally fucked up. So you have to spend another money on injection molded. So that's like 5% out. So we definitely need to keep the error of, I would say, 50% just for all the errors.
So you recommend that for a startup the price for your products should be normally set high?
Yeah, the price of the product has to be premium. That's easier for startups. So if you see Nest, the thermometer, they actually start with I think $299 or $399 when it first start. So the margin is huge. And then they started like $199 a few years ago and now it's like some of the products are $99. And same thing for iPhone. They started with like— iPhone got very famous because every iPhone cost almost $1,000 to $2,000 back in Asia, right? So they start with like everyone with the iPhone can show off with this product and then go to like iPhone 5C or cheaper iPhones. So you mentioned that you're
