| Episodes | 54 |
| Mentions | 75 |
| Cited here | 33 |
| First — last | #35 — #682 |
| Top guests | Jeri Ellsworth, Ahmed, Natasha Baker |
| Related | venture capital |
Startup funding is the acquisition of capital by a new company to finance development, manufacturing, and operation before the business can sustain itself from revenue. For engineering-led and hardware companies, the available routes range from personal savings and friends-and-family cheques through grants, accelerators, and crowdfunding to institutional venture capital, and each route carries distinct consequences for ownership, pace, and risk.[72][109] The structure of the raise determines more than the amount: the size of a round dictates an expected growth rate, the source of the money dictates the obligations attached to it, and the timing of the raise relative to the remaining work frequently determines whether the company survives.[394][163] Because early hardware development can be cheap but production and staffing are not, the gap between a fundable prototype and a funded company is a recurring structural problem in the sector.[94][340]
Sources of early capital
Bootstrapping and personal savings
For a hardware startup that is not doing something unusually capital-intensive, up-front cost is measured in thousands rather than hundreds of thousands of dollars, which puts self-funding within reach of an ordinary engineering salary.[94] A concrete route is to work three or four years after college, bank roughly fifty thousand dollars, and treat that sum as a year of runway instead of raising outside money at the idea stage.[94] Pooled personal savings have a realistic order of magnitude: the three full-time founders of Upverter found that their combined savings bought about four months of company operation before outside money was required.[163]
The case for bootstrapping is control: while the founder is the only funder, the founder owns the whole company and answers to nobody, a position that is lost the moment external money arrives.[72] Having working prototypes in hand before seeking money also puts a hardware startup ahead of the many that pitch on an idea alone, and self-funding those early prototypes is itself read by investors as a positive signal.[147]
Self-funding continues into the early operating period through side income. Before entering an accelerator, Natasha Baker funded SnapEDA by freelancing as a writer for Reuters and Forbes and doing contract coding, which paid the bills but restricted the product to part-time hours.[531]
Friends and family
The first external money for many companies comes from personal networks rather than institutions. Cree co-founder John Edmond recalls that the company’s original capital came entirely from friends and family: “It was basically talking to friends and family, going to a bar and sit there and someone writes you a check for 5,000."<sup><a href="#ref-71" title="Ep 71: An Interview with John Edmond - Luciferous LED Lucubrator">[71]</a></sup> A friends-and-family round is commonly the stepping stone to an institutional seed; one hardware company raised roughly 250,000 from family and friends and then attempted a $2 million seed round, an escalation that frequently fails.[249]
Raising from a large crowd of small backers carries a hidden ongoing cost: a thousand people who each contribute a thousand dollars become a thousand people expecting answers by email.[72]
Revenue-funded growth
A revenue-funded loop for a small hardware product works by buying parts for a batch of roughly a hundred kits, selling them, and reinvesting the proceeds into a larger parts buy; it compounds slowly but never dilutes the founder.[264] The model can reach meaningful scale: a tinkering hobby was grown into a $33 million business with no outside funding at all.[264]
Debt
At the point of starting a company, the founder faces a fork: fund the venture from personal savings or a personal bank loan and retain full ownership while carrying all the downside, or use investor money and shed personal risk at the cost of the company itself.[109] Larry Sears, who built and sold his own hardware business, warns that bank debt is not a substitute for equity at the risky stage, because banks lend only against a guarantee of repayment rather than against the prospects of the venture.[109]
Grants and public programmes
Regional development foundations and civic funds are a real and underused source of startup capital; they are competitive and slow to win but generally attach far fewer strings than equity investors.[79] The obligation attached to such money is usually publicity rather than equity or repayment: the funder wants to showcase the companies it backed as evidence that its programme works.[79] Some national governments compete for startups directly by paying founders’ living expenses to relocate for a fixed period, an arrangement that funds runway without taking equity but requires physically moving the company.[79]
Equity crowdfunding and securities regulation
Before the JOBS Act, United States rules dating from the 1930s restricted private investment to family members and accredited angels, who were required to hold a million dollars in net worth excluding their home; this is why early rounds were historically confined to a small social circle.[87] A ban on general solicitation dating from 1933 likewise meant that for most of the history of startups, fundraising had to happen privately, and a founder’s practical access to investors was limited to whoever was already in their network.[260] The JOBS Act opened a path for a small startup to raise up to a million dollars in ten-thousand-dollar increments from non-wealthy investors, with the company setting the equity terms for each increment.[87]
Accelerators
Startup accelerators exchange equity for a small amount of capital, a fixed programme, and access to investors. Y Combinator’s standard offer at the time was 25,000 in seed funding that was expected to cover the founders’ travel and living expenses on site as well as company costs, so the headline number was not all available for the product.[121]
A batch typically ends in a demo day at which founders give a very short pitch to a room of investors; the actual raise happens in the weeks afterwards, so demo day is the start of the fundraising process rather than its conclusion.[131] The application questions are worth answering even without applying, because they force a founder to state how the business will make money, how equity is split, and what they understand about the market that competitors do not.[268]
Baker judged that the accelerator’s most valuable effect on SnapEDA was not the money itself but that a small amount of funding let her stop consulting and work on the company full time; she also cautions that general accelerators will not teach hardware engineering, and that founders should expect help with scaling and company-building while treating technical instruction as out of scope.[531] On the earliest-stage question of speed, Luke Iseman of Y Combinator holds that before manufacturing and unit cost matter, a hardware startup should not be materially slower than a software one, and that if any step such as populated boards or a printed enclosure adds more than about a week, that is a process problem rather than an inherent property of hardware.[268]
The fundraising process
Fundraising is a full-time occupation, not a side activity. Founders who successfully raise typically report roughly twelve months of full-time effort spent finding someone willing to fund them, and expecting to raise while continuing other full-time commitments is a common misjudgement; a serious raise requires clearing six months or more.[267] The process imposes a structural trade on the founding team: a founder who wants to keep doing engineering cannot also run the raise, so taking charge of the company means the job becomes hunting for money.[267]
Investor fit matters more than investor availability. Finding investors at all is easy; the hard part is finding one who agrees with the intended direction of the product, because a mismatched investor will pull the company somewhere else.[260] A working prototype is the instrument that unlocks funding, and funding in turn is what makes it possible to hire the specialists and support staff needed to make the product good.[407]
Sizing and spending a raise
Runway and unit costs
The practical unit of account when sizing a raise is the engineer-year: a million dollars of funding translates to roughly five people for one year if the company is not also manufacturing.[172] A million dollars is consumed quickly once it covers wages plus production tooling, since injection-mould tooling can run around fifty thousand dollars per mould.[104] Prototype fabrication also scales badly with complexity; quick-turn eight- and ten-layer prototype boards ran about two thousand dollars for a single board.[563]
Even a large raise may not cover production. One hardware company that had raised $15 million needed further capital specifically to fund the bill of materials, since inventory must be paid for before any of it is sold; alongside inventory, engineering salaries are the other dominant line item, so a raise is largely a bet on how many engineer-years the product still needs.[340]
Upverter’s experience illustrates deliberate sizing. The company finished its accelerator with 650,000 of total raised capital before taking its first institutional round.[163]
Capital efficiency
Funded hardware companies frequently keep capital costs down through improvisation. Running Cartesian Co on limited capital, Ariel Briner made a discipline of asking what the cheapest apparatus is that can test the current hypothesis, rather than buying laboratory equipment to do the measurement properly.[260] Building a novel electric propulsion system on startup money, Todd Bailey’s team housed its first hollow cathode in an off-the-shelf plumbing pipe and fabricated a vacuum chamber reaching ten to the minus five torr in house for about fifteen thousand dollars, far below commercial cost, because a founder had the prior experience to build it.[701]
Raising too much
The counterintuitive risk is over-funding. Ben Einstein, an investor specialising in early hardware companies, holds that raising too much money damages a startup more than raising too little.[402] Reflecting on running CastAR, Jeri Ellsworth concluded that the company’s 1.5 million seed supporting a team of about ten was producing steady progress, and that it would have been better served by a deliberate plan for how much to raise rather than taking the largest available round.<sup><a href="#ref-394" title="Ep 394: Jeri Ellsworth and the demise of CastAR (May 28, 2018)">[394]</a></sup> After raising 15 million, CastAR grew from about ten to roughly twenty people in six months—far slower than the raise implied—which set up a mismatch between the capital taken and the pace the company could sustain.[394]
The mechanism is structural: a large round carries an expectation that the company will accelerate and start burning the money quickly, so the size of the raise dictates a growth rate whether or not the product is ready for it, and the venture-funded cycle assumes roughly annual raises with each round sized to be exhausted in about a year—a company that spends more slowly than planned is counted as behind schedule rather than prudent.[394] Ellsworth distinguishes two viable kinds of hardware startup that need different funding: brand-led consumer products that can grow explosively and be sold quickly, and long-horizon deep-technology companies whose payoff is far away.[394]
Failure modes and financial risk
The defining decision point in a funded startup arrives when the cash covers four or five months and the remaining work will take eight; the options at that moment are to pivot, to return to investors, or to wind up and return the money.[163] Upverter’s founders hit this reality check when they compared their bank balance against the work remaining and found they had about one third of the money required, before even accounting for the likelihood of running late.[163]
Companies rarely wind down gradually; the money runs out abruptly, and a company can go from operating normally to having a week of cash and no incoming investment.[351] A distressed company usually cannot admit publicly that it is near bankruptcy, because signalling that position destroys any price it might get in a sale, so silence from a company should not be read as health.[368]
The risk transfers to employees. An engineer who has worked at several startups, Jared Wolff, advises keeping three to five months of personal living expenses in savings before joining one, because layoffs and pay cuts arrive without warning when a round fails to close.[509] Dave Young left a stable instrument-company job for a hardware startup and found it out of money four months after he joined, then continued working there accruing shares instead of a salary.[305] Prospective employees are advised to ask about a startup’s cash flow and runway rather than its headline funding total, since cash on hand says nothing about the rate at which it will be spent; a company with two years of runway will not spend down to zero, and a round of cuts should be expected at the one-year mark and another at six months as management stretches the money.[509]
Easy money is its own failure mode. Recalling his own dot-com-era company, Eric Ries describes an environment in which any plausible-sounding idea attracted a couple of million dollars, funding a product nobody ultimately wanted.[159] At the self-funded extreme, an inventor mortgaged her house and put roughly a quarter of a million dollars of her own money into a product that was already being sold by many others, illustrating why a competitive search should precede any self-funded commitment of that size.[368]
Structural shifts in hardware finance
The financeability of hardware has changed over time. Altera was started in 1983 on $750,000 of secured funding, a benchmark for what it once cost to launch a programmable-logic semiconductor company.[659] Silicon Valley chip companies were once routinely funded to Series C or D before any revenue at all, a pattern that no longer exists and which explains why deep-technology hardware is now much harder to finance.[394] Brett Fox, who raised venture money to build a semiconductor company to scale, reports that even an experienced team with a good reputation found it extremely hard to raise the capital, and judged his to be among the last semiconductor companies that would be venture-backed to scale.[129]
Capital intensity defines the boundary of bootstrapping. Andreas Olofsson, having tried to fund Adapteva’s silicon development from a crowdfunding campaign, concluded that when reaching a shipping product plus a software ecosystem costs millions, bootstrapping is not a realistic option.[254] At the middle of the market, investors looking at consumer electronics and open hardware companies reportedly would not engage below a ten-million-dollar threshold, leaving a gap in which a company is too large for small-company tactics and too small to attract that capital.[127]