How to Bootstrap a Hardware Startup (No VC Playbook)

Short answer: To bootstrap a hardware startup, pick a small, low-BOM product that sells to makers or professionals and does not need regulatory pre-clearance. Prove demand with paid preorders, prototype with free CAD and cheap PCB assembly, design around pre-certified radio modules, and pay for your first small batch with customer money instead of investor money. Then grow each batch only as fast as sales and cash allow.

Hardware is the hard mode of bootstrapping because the cash goes out long before it comes in. Every number below (prices, fees, rules and limits) comes from the vendor's, regulator's or company's own page, checked on 11 Nov 2026. The stage gates, 12-month plan and cash-flow model are ours and are labelled as such.

TL;DR#

• The cash gap is the whole game. You pay for parts, tooling, testing and inventory months before customers receive anything. Every choice below is about shrinking that gap.
• Pick a bootstrap-friendly product. Small enclosure, low bill of materials (BOM), no medical or automotive approvals, sold to people who already buy dev boards, tools and accessories.
• Prototyping is cheap now. KiCad is free under GNU GPL v3. JLCPCB advertises prototype PCBs "From $2.00 / 5 pcs", and its economic assembly lists an $8.18 setup fee and $0.0016 per joint.
• Certify the smart way. In the US a digital device without a radio can usually use the Supplier's Declaration of Conformity. Anything that transmits needs certification, so use a pre-certified module.
• Get paid before you build. Shopify pre-orders can collect full, partial or no payment up front. Kickstarter charges 5% plus processing and requires a working prototype.
• It has been done. Prusa Research, Adafruit and SparkFun all say in their own words that they grew without investors.

Disclosure: ZeroRaised publishes this guide and runs a launch site for bootstrapped products. ZeroRaised has no commercial relationship with any manufacturer, platform or company named here, and no links in this post are affiliate links. Nothing here is legal or compliance advice: check the rules for your product with a test lab or specialist before you sell.

Why hardware is harder to bootstrap than software#

A software founder can build for nothing but time, launch, and get paid the same week. A hardware founder has three costs that arrive before revenue.

1. Inventory. You buy parts and pay a factory before you ship. If a batch does not sell, the cash sits on a shelf.
2. Tooling and setup. Injection moulds, test fixtures, stencils and assembly setup fees are paid once, up front, whatever you sell. Protolabs puts aluminium mould tooling at a "$1,500 minimum" and steel at "$50,000 or more" (Protolabs).
3. Compliance. Most electronics need emissions testing and paperwork before you can legally sell them, and a failed test means a redesign.

Our read: the useful way to think about it is the cash conversion cycle, the number of days between paying a supplier and getting paid by a customer. In software it can be negative (customers prepay annual plans and the hosting bill comes later). In hardware it is often three to six months positive: you pay for parts in month one, build in month two, ship by sea in month three and sell through over the next few months. Venture capital exists in hardware largely to fund that gap. A bootstrapper has to close it another way, and the main tool is customer money: preorders, deposits and crowdfunding.

The other difference is that mistakes are expensive to undo. A software bug is fixed with a deploy. A hardware bug in 500 units can mean a rework, a recall or a lot of refunds. That is why the stage gates later in this guide matter more than they would for an app.

Choose a bootstrap-friendly hardware product#

The product you pick decides whether bootstrapping is realistic at all. Most failed hardware bootstraps chose a product that needed venture-scale money from day one: a consumer robot, a wearable with a custom battery and app, a kitchen appliance on mains power.

Here is our checklist for a product that one person or a small team can fund from sales.

Good bootstrap categories, in our view:

• Dev boards, add-on boards and breakout modules. No enclosure, buyers are technical and forgiving, and the community will tell you what to make.
• Maker kits and tools. Soldering aids, test equipment, keyboard kits, synth modules.
• Accessories for an installed base. Mounts, adapters and upgrades for a popular 3D printer, camera, console or instrument.
• Professional niche devices. A sensor or controller for a specific trade, where buyers pay for time saved and volumes are small but margins are high.

The bootstrapped companies below (Prusa, Adafruit, SparkFun, Pimoroni) all sell to makers and engineers, the most forgiving and vocal buyers in hardware.

Our read: if your idea needs a custom radio, a lithium pack, a companion app and retail distribution, it may still be a good business, but it is probably not a bootstrapped first product. Find a smaller product in the same space that earns the money and the audience to build the big one later.

Real hardware companies that grew without VC#

These examples come from each company's own pages or its founders' own posts, checked on 11 Nov 2026.

Prusa Research#

Prusa's about page says the company "was founded as a one-man startup in 2012 by Josef Prusa, a Czech hobbyist, maker and inventor", and that "There were no investors, no Kickstarter campaigns." The same page says it has grown "into a company of more than 1,200 people" and that every month it designs, builds and ships "over 10,000 printers" (Prusa Research).

Our read: Prusa still sells kits alongside assembled printers (the MK4S kit is cheaper than the assembled version on its store today). Kits are a classic bootstrap move: less labour per unit, a lower price point and buyers who tolerate rough edges.

Adafruit#

Adafruit's about page says it "was founded in 2005 by MIT engineer, Limor "Ladyada" Fried" and is "a 100% woman owned manufacturing company" (Adafruit). Its editorial standards page says the founders "have no investors" and "no loans, no VC (venture capital)" (Adafruit editorial standards).

Our read: a bootstrapped hardware company cannot outspend anyone on ads, but it can out-teach them. Tutorials, build guides and open documentation turn every product into a marketing channel that compounds.

SparkFun#

SparkFun's about page traces it to "our humble beginnings in a Boulder college apartment" in 2003 (SparkFun). In a 2017 post, founder Nathan Seidle wrote: "we've never taken any venture capital or outside investment" and "Because we don't have outside investors, we are able to prioritize a profitable values-driven business" (SparkFun News, 2017). A 2023 post by CEO Glenn Samala describes the company as started by a college student in his dorm room and "completely bootstrapped" (SparkFun News, 2023).

Our read: Seidle's point is that no investors meant the company could choose profit and its own values over growth at any cost. That trade is the core of bootstrapping hardware: you grow slower, but every batch has to pay for the next one.

Pimoroni#

Pimoroni's about page says "Pimoroni Ltd was founded by Jon Williamson and Paul Beech" in 2012, "We're very proud to be an independent manufacturer" and "Pimoroni employs forty people" (Pimoroni). The page does not discuss funding, so we only claim what it says: independent.

Our read: a team of forty is small by hardware standards, and that is the point. You do not need a factory or a hundred staff to be a real, independent hardware manufacturer.

Keyboardio (a crowdfunding case)#

Keyboardio's Model 01 is a useful crowdfunding example, not a "no investors" one, because its pages do not discuss outside funding. Its founders wrote that the campaign hit its "$120,000 funding goal in two and a half hours" (Keyboardio), and a later post listed "Backers: 2073 Total dollars pledged: $652,001" and "Percentage of original goal reached: 543%" (Keyboardio). The same post notes 608 backers were already on their mailing list.

Then production reality hit. In a May 2017 update they wrote that the plan was for Jesse to oversee manufacture of the first 100 keyboards in Shenzhen in April, and "That didn't happen and we're pretty bummed out about it." On emissions testing: "We almost passed" and "The Model 01 had just a little bit too much electromagnetic radiation in the 35-115Mhz range" (Keyboardio).

Lesson: two in one. A mailing list built before launch carried the campaign. And a small EMC failure can push back delivery, so test early, which is covered below.

The stage-gate plan (ours)#

Hardware punishes skipping steps. This is the gate structure we would use, with a test to pass before you spend the next tranche of money.

EVT, DVT and PVT (engineering, design and production validation test) are common industry shorthand. You do not need the formality, but you do need the discipline of not ordering 500 of something you have built 5 of.

Validate with preorders and deposits#

Interest is cheap. Money is the signal. Before you spend on an engineering prototype, ask for payment.

If you are new to this, our guide on how to validate a SaaS idea before building covers interviews and smoke tests that apply just as well to hardware. Pair it with a waitlist before launch to collect emails from people not ready to pay yet. Keyboardio's 608 pre-subscribed backers show why that list matters.

The hardware-specific part is how you take money for something that does not exist yet.

Option 1: Full-price preorders on your own store#

Shopify's help centre says: "You can sell products that are out of stock or not available yet by setting them up as pre-orders." It adds that, depending on the options you offer, "you can collect full, partial, or no payment" up front "and then you can fulfill and collect the remaining payment at a later date." Two conditions: "To use pre-orders, you need to install a pre-order app", and pre-orders "are currently only available to merchants using Shopify Payments or Paypal Express" (Shopify Help Center).

Option 2: Save the card now, charge later (Stripe)#

If you build your own checkout, Stripe's docs say: "The Setup Intents API lets you save a customer's payment details without an initial payment," so you can "charge them in the future" (Stripe docs). This suits a "reserve now, pay when it ships" offer.

Do not plan on a card hold instead. Stripe says "Usually, an authorization for an online card payment is valid for 7 days," and if it expires before capture, "the funds are released" (Stripe docs). Hardware timelines are far longer than a week. In the US, Stripe's standard card rate is "2.9% + 0.30 USD" (Stripe support).

Which to choose#

Our read:

• Charge in full if you are confident on timing and want the cash to fund production. This is what closes the cash gap. It also means you owe refunds if you slip badly, so be honest about dates.
• Take a deposit (say 10% to 30%) if you want a strong signal without holding too much customer money.
• Save cards, charge on shipping if you do not need the cash and want to minimise refund risk. It is the weakest signal, because some saved cards will fail or be cancelled by the time you charge.

Whichever you pick, write a clear policy on the product page: estimated ship month, what happens if you are late, and how refunds work. Consumer protection rules for pre-sales vary by country, so check yours.

Crowdfunding a hardware product#

Crowdfunding is preorders with an audience attached. Our guide to crowdfunding a software product covers campaign pages, reward tiers and pre-launch lists. Here we cover only what is different for hardware: the platform rules, the fees and the delivery problem.

Platforms and fees#

Kickstarter's own guide is blunt about hardware: building a prototype that shows off all your product's core features "is not only required by Kickstarter's rules", "any major feature you're promising to backers will need to be demonstrated fully on your page", "realistic renderings are not allowed", and "Without a good prototype, you are too early for Kickstarter" (Kickstarter).

Indiegogo now belongs to Gamefound; Gamefound's blog headline reads "Gamefound acquires Indiegogo" (Gamefound).

Hardware-specific crowdfunding traps#

Our read, from watching many hardware campaigns:

1. Pricing for backers too low. Early-bird pricing that only works if the BOM quote holds, the factory hits yield and shipping costs do not move. Price backer tiers off a real manufacturer quote plus a contingency, not your prototype costs.
2. Shipping and duties. International shipping, import VAT and duties can eat the margin on a cheap product. Decide in advance whether backers pay shipping separately (many campaigns collect it after the campaign) and which countries you ship to.
3. Over-funding. A campaign that raises five times its goal sounds like a win. It also means five times the units, possibly a different manufacturing process, and a supply chain that has never been tested at that size. Keyboardio hit 543% of its goal and later delays followed.
4. Stretch goals that add features. Every stretch goal is a new engineering and compliance risk. Make stretch goals colours or accessories, not new radios.
5. Delivery dates. Add your expected time for certification, a failed test, Chinese New Year if you build in China, and sea freight. Then add more.

Crowdfunding is one of several no-VC funding routes; our ranking of ways to fund a product without venture capital compares it with grants, loans and revenue-based options.

Your prototyping stack, with current prices#

Prototyping used to be the expensive part. It no longer is. Prices below were shown on each vendor's site on 11 Nov 2026 and change often, so treat them as a snapshot.

Electronics design: KiCad (free)#

KiCad describes itself as "an open source software suite for Electronic Design Automation (EDA)" that runs on Windows, Linux and macOS and "is licensed under GNU GPL v3" (KiCad). Every major PCB fab accepts KiCad output. There is no reason to pay for EDA software at this stage.

PCBs and assembly#

JLCPCB notes that PCB "Costs depend on size, layer count, surface finish, and quantity", so the headline price is for small, simple boards. On assembly, extended (less common) parts carry extra labour fees, and economic assembly takes 2 to 50 boards with a build time of 1 to 3 days, while standard assembly runs from 2 up to 80,000 boards (JLCPCB capabilities).

Worked example (illustrative): a small board with 300 SMT solder joints, 10 boards via JLCPCB economic assembly. Joints: 300 x 10 x $0.0016 = $4.80. Add the $8.18 setup fee and $1.53 stencil and the assembly labour is about $14.51 for the batch, before parts, the bare boards, extended-part fees and shipping. The point is that assembly labour is no longer the barrier. Parts and your time are.

Enclosures and mechanical parts#

A desktop 3D printer covers most early enclosure work.

Prices are as displayed in dollars on Prusa's store when we checked; your region and taxes may change them. Bambu Lab is the other popular choice, but its store did not load for us today, so we have not quoted its prices.

If you do not want a printer, outsource parts. JLCPCB lists 3D printing "From $0.30" and CNC machining "From $5.00" (JLCPCB); PCBWay lists 3D printing "From $4.98" and CNC "From $25" (PCBWay). Xometry says you can "Upload your files, get an instant quote, and order online", with "no minimums required" (Xometry). Protolabs says its process can "produce parts in as fast as 1 day" (Protolabs).

Our read: for a first product, a 3D printed or off-the-shelf enclosure is fine. Many maker products ship as bare boards with an optional case. An enclosure that looks hand-made is a smaller problem than a mould you paid for and then had to change.

Design for manufacturing (DFM) before you order a batch#

DFM means designing so that the factory can build your product reliably and cheaply. A prototype that works on your bench can still be expensive or unreliable at 500 units. Our checklist:

• Use common parts. At JLCPCB, "extended" parts carry an extra labour fee (JLCPCB). Basic parts are cheaper and more likely to be in stock.
• Avoid single-source parts where you can, and check stock levels before committing. Pick parts with drop-in alternatives.
• Keep everything on one side of the board if you can. JLCPCB's standard assembly setup is $25.56 single-sided vs $51.12 double-sided (JLCPCB).
• Design in test points and a simple test procedure (or a test jig) so every unit can be checked before it ships.
• Program firmware at the factory or with a fast jig. Flashing 500 boards by hand is a week of your life.
• Ask for a DFM review. Seeed lists a "Free DFA/DFM Review" with Fusion orders (Seeed Fusion), and Protolabs says it sends "design for manufacturability (DFM) analysis and real-time pricing" with quotes (Protolabs).
• Design the packaging early. Size and weight drive shipping costs, and shipping often decides whether a cheap product is profitable.

Small-batch vs injection-mould tooling#

The big mechanical decision is when to pay for an injection mould. Moulded parts are cheap per unit, but the mould is a large fixed cost.

What the vendors say on their own pages:

• Protolabs: aluminium tooling "$1,500 minimum (depending on capacity)", steel "$50,000 or more" (Protolabs blog, Dec 2025). Its help centre says injection-moulding "Prices start around $1,495, depending on part geometry and complexity" (Protolabs help).
• Protolabs on-demand moulding has "No minimum order quantity". Its prototype moulds are "guaranteed for at least 2,000 shots" (Protolabs on-demand).
• PCBWay says it can make "molds ready in 15 days" and lets you "order as little as a single prototype up to 10,000 production parts" (PCBWay injection moulding).
• Xometry says T1 samples arrive "As fast as 5 business days; 3 weeks is typical for most projects", and that production quantities and tooling need "manual review" (Xometry).

Our read: treat these minimums as floors. A real enclosure with two halves, snap fits and a button can need more than one mould and cost several times the minimum. Get quotes on your actual part. As a rule of thumb, stay with 3D printing, an off-the-shelf enclosure or no enclosure until a batch sells out and the design has stopped changing. Then a mould is an investment in margin rather than a bet.

Certification: FCC, CE, UKCA and friends#

This is where bootstrapped hardware founders most often get surprised. We quote the rules; we are not your compliance engineer. Get a quote from an accredited test lab early, and do a cheap pre-compliance scan before formal testing.

United States: FCC Part 15#

The key distinction is between intentional and unintentional radiators.

• An intentional radiator is "A device that intentionally generates and emits radio frequency energy by radiation or induction" (Wi-Fi, Bluetooth, LoRa and so on).
• An unintentional radiator generates radio frequency energy "for use within the device" but is "not intended to emit RF energy by radiation or induction."
• A digital device is "An unintentional radiator (device or system) that generates and uses timing signals or pulses at a rate in excess of 9,000 pulses (cycles) per second" (47 CFR §15.3). Almost anything with a microcontroller qualifies.

Unintentional radiators "shall be authorized prior to the initiation of marketing", and the §15.101 table lists "Other Class B digital devices & peripherals" as "SDoC or Certification" (47 CFR §15.101). Intentional radiators are different: "all intentional radiators operating under the provisions of this part shall be certified", subject to listed exceptions (47 CFR §15.201).

The two routes:

Use pre-certified radio modules#

If your product needs Wi-Fi or Bluetooth, the bootstrap answer is almost always a pre-certified module rather than a custom radio design. Under §15.212, a single modular transmitter is "a completely self-contained radiofrequency transmitter device" that can be certified on its own. If the module's FCC ID is not visible once installed, "the outside of the device into which the module is installed must also display a label referring to the enclosed module", for example "Contains FCC ID: XYZMODEL1" (47 CFR §15.212).

The FCC's detailed module guidance lives in KDB 996369, "Modules, Module Certification, 15.212", which includes a Module Integration Guide for host manufacturers (FCC KDB 996369). Read it and the module vendor's integration notes before you lay out the board: antenna placement and keep-out areas usually have to match the certified design.

Our read: the module does not make your product exempt. The rest of your device is still a digital device and an unintentional radiator, so you still need to show it complies. But you avoid the most expensive and slowest part, certifying a new transmitter.

European Union: CE marking#

The EU's Your Europe portal says that with the CE marking "the manufacturer indicates that a product meets the requirements set out in EU product rules." For most electronics the manufacturer self-assesses (a notified body is needed only where the law requires one), compiles a technical file kept for at least 10 years, and signs an EU declaration of conformity (Your Europe). The Commission adds that "a CE marking does not indicate that a product have been approved as safe by the EU" (European Commission).

The directives most small electronics founders meet:

• EMC Directive 2014/30/EU "limits electromagnetic emissions from equipment" and "also governs the immunity of such equipment to interference" (EC).
• Radio Equipment Directive 2014/53/EU for anything with a radio (EC). The Commission says its cybersecurity delegated act, Regulation (EU) 2022/30, "entered into application in August 2025", and that it is being repealed by Delegated Regulation (EU) 2026/339 when the Cyber Resilience Act applies; "The main obligations introduced by the CRA will apply as of 11 December 2027."
• Low Voltage Directive 2014/35/EU covers equipment between "50 and 1000 V for alternating current" and "75 and 1500 V for direct current" (EC). A USB-powered gadget sits below those thresholds, though the external power supply you bundle does not.
• RoHS "currently restricts the use of ten substances", including lead, cadmium and mercury (EC RoHS). Order RoHS-compliant parts and keep supplier declarations.
• WEEE "Requires the separate collection and proper treatment of WEEE" (waste electrical and electronic equipment), with national producer registers (EC WEEE). Check each country's registration rules before you ship there.
• General Product Safety Regulation (EU) 2023/988 "is applicable since 13 December 2024" (EU-OSHA). It matters if you sell consumer products into the EU from outside it; check what it requires of you before listing there.

United Kingdom: UKCA or CE#

GOV.UK says "the UK continues to recognise the CE marking, alongside or in place of the UKCA marking, for the Great Britain market" (GOV.UK). For radio equipment specifically, the guidance says the regulations "extend recognition of CE marking indefinitely in GB" and that "The manufacturer must keep technical documentation and the declaration of conformity for 10 years" (GOV.UK Radio Equipment Regulations 2017). For a small founder this means one set of CE paperwork can cover both markets in many cases.

How to keep compliance affordable#

Our read:

1. Decide your markets first. US-only for batch one halves the paperwork.
2. Use a pre-certified module for any radio, and follow its reference layout.
3. Design for EMC from the first PCB: solid ground plane, short high-speed traces, filtering on cables.
4. Book a pre-compliance scan at a local lab or makerspace before formal testing. Keyboardio's near-miss in the 35 to 115 MHz range is the kind of failure a pre-scan catches.
5. Get formal test quotes from two or three accredited labs, and budget for one retest.
6. Avoid batteries in the first product if you can. Lithium cells bring shipping and safety rules, and the EU Batteries Regulation (EU) 2023/1542 has applied since 18 February 2024 (EC Batteries).

Manufacturing partners and minimum order quantities#

For a first product you will usually choose between:

• Online PCB assembly services (JLCPCB, PCBWay, Seeed Fusion). Low minimums: JLCPCB economic assembly takes 2 to 50 boards, PCBWay quotes 1 to 20 boards, Seeed says "No minimum order requirements". Ideal for batches up to a few hundred if your product is mostly a PCB.
• Online manufacturing networks (Protolabs, Xometry) for machined, printed and moulded parts with no minimum order quantity on the services quoted above.
• A contract manufacturer (CM) that does assembly, test, packaging and sometimes sourcing. Better for complex products, but minimums and engineering fees are quoted per project, so you will need to ask.
• Assembling in-house. Many maker companies start by hand-assembling final products from factory-built boards. It is slow, but it teaches you exactly where the product is hard to build.

Our read: at batch one, keep as many parts of the build with online services as you can. You get public prices, small minimums and fast feedback. Move to a CM when volume or complexity makes it worth paying their engineering fees, and get at least two quotes.

Pricing and margin: from BOM to retail#

Hardware pricing starts from the landed cost per unit: BOM (parts plus PCB plus assembly), enclosure, packaging, inbound freight, duties and test time. Then add your sales channel costs.

Worked example (illustrative, our numbers)#

A USB-powered sensor board in a small 3D printed case, sold to makers.

Now the same unit through other channels (still illustrative):

• Wholesale to a retailer. A common retail rule of thumb is that the retailer roughly doubles your wholesale price (keystone pricing). At $59 retail you would sell at around $29.50, leaving about $8 over landed cost before your own overheads. That is why many bootstrapped hardware brands sell direct first.
• Tindie. Tindie says it "charges a 5% marketplace fee on each paid order", with payment processing charged separately (Tindie). On $59 that is $2.95 plus processing.
• Amazon. The Professional plan is "$39.99/month + selling fees" and the referral fee for Consumer Electronics is 8% (Amazon), so $4.72 on $59, plus Fulfillment by Amazon fees if you use FBA. Amazon publishes those via its revenue calculator rather than a simple table, so run your product's size and weight through it.

Our read: aim for a landed cost of no more than a third of your direct retail price on the first product. That leaves room for processing, shipping subsidies, returns, a later wholesale channel and the mistakes you will make on batch one. If the BOM forces a retail price your buyers will not pay, change the product, not the spreadsheet.

Worked cash-flow example: the first 500-unit run (illustrative)#

This is our model, not any company's real numbers. It uses the per-unit example above, the Protolabs minimum aluminium tooling figure as a placeholder, and assumptions that are clearly marked.

Assumptions: founder puts in $5,000 of savings. 200 units presold at $59, charged in full. Manufacturer wants 50% deposit and 50% before shipping (common, but terms vary). Compliance testing budget $3,000 (placeholder; get lab quotes). One aluminium mould at $1,500 (Protolabs' stated minimum; your quote may be higher). Customers pay shipping separately. Remaining 300 units sell over three months after stock arrives.

Totals: revenue $29,500; processing $1,005; product cost $10,750; one-off costs $4,500. The run ends with $18,245, $13,245 more than the founder put in.

Three things to notice:

1. The low point is month 3, at $1,148. One failed compliance test, a parts price rise or a freight surprise and you are out of money with 500 units in a warehouse. That is why preorders matter so much: without the 200 presold units, this plan does not work on $5,000 of savings.
2. Most of the profit is in the units you have not sold yet. Your batch-two budget depends on sell-through of batch one.
3. Batch two is easier. The mould and certification are paid. If nothing changes in the design, the next 500 units cost $10,750 plus freight, and you can fund them from cash.

Funding the gap without VC#

If preorders and savings do not cover the low point, here are the no-VC options in roughly the order we would try them:

1. Customer money. Preorders, deposits, crowdfunding, or a distributor or large customer paying up front for a batch.
2. Grants. In the US, SBIR, and in the UK, Innovate UK, can fund deep-tech hardware R&D. Our startup grants guide covers eligibility and how competitive they are.
3. SBA-backed loans (US). The SBA says "The microloan program provides loans up to $50,000" (SBA microloans), and "The maximum loan amount for a 7(a) loan is $5 million", with SBA Express loans up to $500,000 (SBA 7(a), SBA 7(a) types). Lenders decide eligibility; a first-time hardware founder without revenue may struggle to qualify.
4. Inventory and purchase-order financing. Some lenders finance a confirmed purchase order or stock against future sales. Terms vary widely, so compare the total cost to the margin on the batch.
5. Revenue-based financing. Mostly built for recurring revenue, which hardware rarely has, but some providers fund e-commerce inventory. Our revenue-based financing guide explains how repayments work.

Our read: debt on inventory is only safe once you have proof of sell-through. Borrowing to make your first batch is borrowing against a guess.

Fulfilment: getting units to customers#

Options, from cheapest to most outsourced:

• Ship it yourself. For the first few hundred units, packing at home or in a small unit is the cheapest option and teaches you what breaks in transit.
• Third-party logistics (3PL). ShipBob does not publish prices; its pricing page says "All quotes are customized for each customer" (ShipBob). Ask several 3PLs for quotes based on your size, weight and monthly orders.
• Fulfillment by Amazon. Useful if Amazon is a major channel; fees come on top of the referral fee above.

Our read: stay in-house until packing is eating more of your week than product work. Then move to a 3PL.

The 12-month plan (ours)#

Many products take longer than this, especially if certification needs a redesign. Treat it as an order of operations, not a promise.

Common mistakes#

1. Choosing a product that needs VC. Custom radios, batteries, apps and retail distribution in product one.
2. Ordering inventory before preorders. Your first batch should be mostly sold before it exists.
3. Pricing off prototype costs. Price off a manufacturer quote plus contingency, including shipping and processing.
4. Skipping pre-compliance. Finding an EMC problem after the batch is built is the most expensive way to find it.
5. Paying for a mould too early. Until the design stops changing, printed or off-the-shelf enclosures are cheaper.
6. Ignoring the cash low point. Model month by month, not just total profit.
7. Over-funding a crowdfunding campaign without a plan to scale. More units can mean a different process and new problems.
8. No content engine. Hardware buyers search, read and watch before they buy. Tutorials and build guides are a moat; build yours early.

Launch checklist#

• Product passes the bootstrap-friendly checklist
• 20+ buyer conversations logged
• Working prototype in users' hands
• Paid preorders or deposits hit the target you set
• Costed BOM; landed cost no more than about a third of retail (our rule of thumb)
• DFM review done; test procedure written
• Radio via pre-certified module, labelled correctly
• Pre-compliance scan passed; formal testing booked
• Declarations, technical file and labels ready for each market
• Manufacturer quote and payment terms in writing
• Month-by-month cash plan with a positive low point
• Shipping, duties and returns policy published
• Launch plan, using our product launch checklist for bootstrapped makers

Frequently asked questions#

Can you really bootstrap a hardware startup without VC?#

Yes, if you pick the right product. Prusa Research's about page says it began as a one-man startup in 2012 with no investors, Adafruit's editorial standards page says its founders have no investors and no VC, and SparkFun's founder wrote in 2017 that it never took venture capital or outside investment. The pattern is a small, high-margin product sold to makers or professionals, demand proven before inventory is bought, and batches funded from sales.

How much does it cost to prototype a hardware product?#

For a simple electronic product, prototype costs can be low. KiCad is free and open source. JLCPCB advertises prototype PCBs from $2 for 5 boards (now tied to its JLCONE desktop app), and its economic assembly service lists an $8.18 setup fee, a $1.53 stencil and $0.0016 per solder joint for 2 to 50 boards. A Prusa MK4S printer is $925 assembled on Prusa's store. Certification testing, moulds and a production batch are the real costs, and you should get quotes for those.

Do I need FCC certification for my hardware product?#

Almost every electronic product sold in the US needs FCC equipment authorization before marketing, but the route differs. A digital device without a radio is an unintentional radiator and Part 15 lets most Class B digital devices use either the Supplier's Declaration of Conformity (SDoC) or certification. A device that transmits is an intentional radiator and must be certified. Using a pre-certified radio module means the radio is already certified, though your finished product still has to comply and carry a "Contains FCC ID" label.

Should I launch my hardware product on Kickstarter?#

Only once you have a working prototype. Kickstarter says a prototype showing all core features is required, and that realistic renderings are not allowed. It charges a 5% fee on funded projects plus payment processing of roughly 3 to 5%. Crowdfunding works best when you already have a mailing list, a costed bill of materials, a manufacturer quote and a fulfilment plan, because the hard part is delivering, not raising.

How do I fund the first production run without investors?#

Use customer money first: paid preorders on your own store, a crowdfunding campaign, or deposits from a distributor. Then layer in cheap debt if needed: the SBA microloan program provides loans up to $50,000 and 7(a) loans go up to $5 million. Keep the first batch small enough that you can pay for it from preorders plus savings you can afford to lose, and grow batch sizes as sell-through proves demand.

Launch your hardware on ZeroRaised#

ZeroRaised is a launch directory for products built without institutional VC, and that includes boards, kits and devices. When your preorder page or first batch is ready, submit your product for free. For more no-VC playbooks, join the ZeroRaised newsletter.

Sources#

• KiCad: About, last checked 11 Nov 2026
• JLCPCB homepage, last checked 11 Nov 2026
• JLCPCB PCB assembly price, last checked 11 Nov 2026
• JLCPCB PCB assembly capabilities, last checked 11 Nov 2026
• PCBWay homepage, last checked 11 Nov 2026
• PCBWay PCB assembly, last checked 11 Nov 2026
• PCBWay injection moulding, last checked 11 Nov 2026
• Seeed Studio Fusion PCB assembly, last checked 11 Nov 2026
• Seeed Studio Fusion PCB, last checked 11 Nov 2026
• Prusa MK4S, last checked 11 Nov 2026
• Prusa CORE One, last checked 11 Nov 2026
• Protolabs homepage, last checked 11 Nov 2026
• Protolabs pricing and payment options, last checked 11 Nov 2026
• Protolabs: aluminum vs steel tooling, last checked 11 Nov 2026
• Protolabs on-demand manufacturing, last checked 11 Nov 2026
• Xometry homepage, last checked 11 Nov 2026
• Xometry injection molding service, last checked 11 Nov 2026
• 47 CFR §15.3 definitions, last checked 11 Nov 2026
• 47 CFR §15.101, last checked 11 Nov 2026
• 47 CFR §15.201, last checked 11 Nov 2026
• 47 CFR §15.212, last checked 11 Nov 2026
• 47 CFR §2.906, last checked 11 Nov 2026
• 47 CFR §2.907, last checked 11 Nov 2026
• FCC KDB 996369, last checked 11 Nov 2026
• Your Europe: CE marking, last checked 11 Nov 2026
• European Commission: CE marking, last checked 11 Nov 2026
• European Commission: EMC Directive, last checked 11 Nov 2026
• European Commission: Radio Equipment Directive, last checked 11 Nov 2026
• European Commission: Low Voltage Directive, last checked 11 Nov 2026
• European Commission: RoHS Directive, last checked 11 Nov 2026
• European Commission: WEEE, last checked 11 Nov 2026
• European Commission: Batteries, last checked 11 Nov 2026
• EU-OSHA: General Product Safety Regulation, last checked 11 Nov 2026
• GOV.UK: Using the UKCA marking, last checked 11 Nov 2026
• GOV.UK: Radio Equipment Regulations 2017 (Great Britain), last checked 11 Nov 2026
• Shopify Help Center: pre-orders, last checked 11 Nov 2026
• Stripe docs: save and reuse payment methods, last checked 11 Nov 2026
• Stripe docs: place a hold on a payment method, last checked 11 Nov 2026
• Stripe support: US card pricing, last checked 11 Nov 2026
• Kickstarter fees guide, last checked 11 Nov 2026
• Kickstarter: creating your prototype and manufacturing plan, last checked 11 Nov 2026
• Indiegogo: what changed with the platform upgrade, last checked 11 Nov 2026
• Gamefound: Gamefound acquires Indiegogo, last checked 11 Nov 2026
• Crowd Supply: how are funds received, last checked 11 Nov 2026
• Tindie: sell on Tindie, last checked 11 Nov 2026
• Amazon: selling plans and fees, last checked 11 Nov 2026
• ShipBob pricing, last checked 11 Nov 2026
• SBA 7(a) loans, last checked 11 Nov 2026
• SBA types of 7(a) loans, last checked 11 Nov 2026
• SBA microloans, last checked 11 Nov 2026
• Prusa Research: About us, last checked 11 Nov 2026
• Adafruit: About, last checked 11 Nov 2026
• Adafruit: Editorial standards, last checked 11 Nov 2026
• SparkFun: About, last checked 11 Nov 2026
• SparkFun News: Colorado doesn't need to replicate Silicon Valley (2017), last checked 11 Nov 2026
• SparkFun News: Finding balance and the road ahead (2023), last checked 11 Nov 2026
• Pimoroni: About us, last checked 11 Nov 2026
• Keyboardio: Model 01 Kickstarter, 12 hours, last checked 11 Nov 2026
• Keyboardio: Day 32 stats, last checked 11 Nov 2026
• Keyboardio: May 2017 update, last checked 11 Nov 2026