Your first hardware customer is the first organization that gives something up to get your product. Money, a budget owner's name, an installation slot, an engineer's month. Interest costs a buyer nothing, and a pipeline full of it tells you nothing.
Most hardware startups do not lack interested buyers. They lack a way to tell which of those buyers has committed. A lab that loved the demo, a strategic that signed a letter of intent, a plant manager who wants a pilot: all three feel like customers in the board update. Only one of them may have done anything that costs them.
This note is about the first five to ten accounts. Which commitments count, where to look for buyers who can make them, and how to write the first deal so it leads to the second.
Why the first customer is harder in hardware.
A software buyer can try a product on a credit card and cancel next month. A hardware buyer has to find floor space, a power drop, a safety review, an integration window and someone to run the machine. Each of those has an owner, and each owner can stop the deal.
The numbers make the point. A typical B2B buying group runs 6.3 to 6.8 people (Gartner and 6sense composite). A manufacturing technology deal averages a 124-day cycle at a 19 percent win rate (Digital Bloom), and serious capital equipment commonly runs 9 to 18 months. The median hardware pilot converts to a production contract 12 percent of the time (IDC, 2025), roughly one in eight. All four sit on the Hardware GTM Benchmarks 2026 page with their sources.
Put those together and the first-customer problem changes shape. You are not looking for someone who likes the product. You are looking for a buyer whose organization will fund, install and operate it, and you need to know that before you spend a quarter of runway on the account.
Rank every early commitment by what it costs the buyer.
The useful question about any early signal is simple: what did the buyer give up to make it? The more it cost them, the more it tells you. Here is the order I use, from weakest to strongest.
| Commitment | What the buyer gave up | What it proves | What it does not prove |
|---|---|---|---|
| Interest | An hour. A demo, a site visit, an enthusiastic email. | The problem is real to at least one person. | That anyone with a budget cares, or that the problem ranks high enough to fund. |
| Letter of intent | A signature on a document that often states it is not binding. | Someone senior will put their name near yours. | Price, quantity, date or a budget owner, unless the letter names them. |
| Paid pilot | Money from a named budget, plus site access and staff time. | The buyer will spend to learn. If the production path is written in, it proves much more. | Production. At 12 percent median conversion (IDC, 2025), the pilot alone is not the sale. |
| Conditional order | A purchase order or contract that triggers on criteria you both wrote down. | Budget, price and decision rights exist. The buyer has named what "good enough" means. | That you will meet the criteria, or that the date holds if the site slips. |
| Binding order or deposit | Cash or an unconditional obligation to take the product. | Everything above, plus priority inside the buyer's organization. | That you can deliver. A signed order you cannot build turns into a reference against you. |
Two things about that table matter more than the rows.
First, the ranking is about the buyer's cost, not your effort. A letter of intent can take months to negotiate and still commit the other side to nothing. The contract vehicle note walks through the same trap at a larger scale: an agreement that grants access and obligates no spending, reported as if it were demand.
Second, every row can be upgraded. An LOI that names a quantity, a price band and the production budget owner is worth far more than one that names a partnership. A pilot that writes the production terms in at signature is a different instrument from a pilot that ends in a report. The work of the first year is moving each account up one row.
Where the first customers actually come from.
Founders usually start with the buyers who are easiest to reach: innovation groups, corporate venture arms, R&D labs. Those teams take meetings, run evaluations and sign pilots. They rarely own the budget that pays for production, and that handoff is where the one-in-eight number comes from.
Look instead for three conditions in the same account.
- A budget line that already exists. The buyer is already paying for the problem, in labour, scrap, downtime, outsourced work or a service contract. You are asking them to move money, not find new money. The budget line note covers how to find it.
- An operator who owns the pain. A plant manager, a fleet lead, a site director. Someone whose own number gets better if your product works, and who can get it installed.
- A decision path you can name. Who signs, who can veto, and what each of them needs to see. With a 6.3 to 6.8-person committee, a deal with one contact is a deal with five unknowns. The buying committee note maps the usual seats.
An account with all three is worth more than ten accounts with a champion and no budget. That is also why early customer lists in hardware are short. A startup that can name eight accounts where all three conditions hold has a better first year ahead of it than one that can name eighty logos.
Write the first deal so it leads to the second.
The first deal sets the pattern for every deal after it, so structure matters more than size. Four terms do most of the work.
1. The buyer pays for the pilot.
A free pilot tests your product. A paid pilot tests the buyer. It proves a budget exists, it puts a finance person in the loop early, and it makes the buyer's own team care whether the pilot succeeds. The amount matters less than where it comes from. If the pilot money comes from the same budget that would fund the rollout, you have already crossed the handoff that kills most pilots.
2. Success criteria are written before installation.
Agree in writing what the pilot has to show, measured how, by whom and by when. Use the buyer's units: hours saved, parts per shift, uptime, cost per unit of output. A pilot that ends with "promising results" and no agreed test is a pilot that restarts at the next budget cycle.
3. The production path is named at signature.
Before the pilot starts, name the production budget owner, the approval the rollout needs, the indicative commercial terms and the date the decision will be made. This is the single change that moves a pilot toward production. The pilot to production pillar goes deeper on it.
4. The buyer's team does real work.
An operator who spends time learning your system has a reason to want it kept. Ask for named people, a training plan and a site owner. A pilot the buyer's team never touched is easy to end.
If the pilot met every criterion tomorrow, who signs the production order, from which budget, and when? If nobody on the buyer's side can answer, the deal is interest with a longer name.
Who sells the first ten.
The founder does. The first customers are buying a relationship with the people who built the product as much as they are buying the product. They need to hear the technical limits from someone who knows them, and they need to believe the company will still be answering the phone in two years.
Founder-led selling at this stage is also how you learn what the sales job actually is. Which objections repeat, which roles block, how long each stage really takes, and which commitment level each type of buyer will make. Hire before you know those answers and you hire against a guess. The first sales hire note covers what to have written down before the posting, and the hiring pillar covers the sequence after it.
What the founder should not do is carry the whole motion alone past the point where it repeats. Once three or four accounts have moved up the table the same way, the motion can be written down and handed to someone else.
Count commitments, not conversations.
The forecast is where early-stage hardware companies most often fool themselves, because the board wants a number and every conversation feels like progress. Keep it honest with three rules.
- Report accounts by the row they have reached. Interest, LOI, paid pilot, conditional order, binding order. Show movement between rows each month. That tells the board more than a weighted pipeline total.
- Never roll up a figure the buyer did not commit. An LOI's headline value, a framework ceiling and a pilot's "potential rollout" do not belong in revenue forecasts. Keep them in a separate column.
- Record the next buyer commitment and its date. Not your next step. Theirs. If an account has no next commitment with a date, it is not moving, whatever the last meeting felt like.
The arithmetic follows. At a 19 percent win rate, required pipeline coverage works out to roughly 5.3x (derived on the benchmarks page). That multiple only means something if the pipeline is counted in commitments. Count conversations and any coverage number is fiction. The sales cycle note covers how to plan the calendar once the commitments are real.
What to do this week.
- List every account you call a customer or a near-customer. Put each one in a row of the table above. Be strict.
- For each account, write the one commitment that would move it up a row, and who on the buyer's side would have to make it.
- For any pilot in flight, answer the production question: who signs, from which budget, by when. If you cannot, that is the next conversation.
- Drop, or deprioritize, accounts where no budget line, operator owner or decision path exists. Put the time into the accounts where they do.
A working business case helps with the second and third steps. The hardware business case template is a seven-field brief a champion can take into their own approval meeting.
Frequently asked questions
How does a hardware startup get its first customers?
Start with accounts that already pay for the problem, have an operator who owns it and have a decision path you can name. Sell the first deals founder-led, charge for the pilot, write success criteria before installation and name the production budget owner at signature.
Is a letter of intent a customer?
Usually not. A letter of intent is commonly written as non-binding, and many name no price, quantity, date or budget owner. Treat an LOI as a signal that a senior person is willing to be associated with you, then work to convert it into a paid pilot or conditional order.
Should a hardware startup offer a free pilot?
A paid pilot is stronger. Payment proves a budget exists, involves finance early and gives the buyer's team a reason to make the pilot succeed. If the pilot is paid from the budget that would fund the rollout, it is stronger still.
How many first customers does a hardware startup need before hiring sales?
There is no fixed count. Hire when several accounts have moved from pilot to order the same way, so the motion can be written down. Until then the founder is still learning what the sales job is.
What should a hardware startup report to its board about early customers?
Report accounts by commitment level (interest, letter of intent, paid pilot, conditional order, binding order), the movement between levels and each account's next buyer commitment with a date. Keep non-binding headline values out of revenue forecasts.
Bring your customer list to the table.
If you are not sure which of your early accounts has committed, that is the question to answer first. The Hardware Go-to-Market Diagnostic reviews the buyer, business case, pipeline, team and founder dependence. A Signal Audit is a focused starting point for one account. For the wider picture of how SignalForge works with hardware and deep tech companies, start with GTM consulting for deep tech and hardware or go-to-market consulting for hardware startups.