Why the First Pilot Decides More Than the First Sale

No business buys unproven hardware at scale. The pilot is the mechanism by which a company de-risks you, and it is also the only place you will get honest data about how your product behaves in an environment you did not control. A pilot that runs well produces three assets at once: a reference customer, a documented performance result, and a list of the twelve things your product does wrong that no lab test would have found.

It also has a failure mode that founders walk into repeatedly. A pilot that drags on with no defined endpoint, no success criteria, and no price attached is not a sale in progress. It is free consulting, and it can consume a year of a small company's engineering capacity.

Picking the Right Customer, Not the Biggest One

The instinct is to chase the largest logo that returns a call, and for a first pilot that is usually wrong. A Fortune 500 pilot means procurement, security review, vendor onboarding portals, and a six-month approval cycle before anyone plugs anything in.

Better criteria for the first one:

  • The pain is quantified. They can tell you what the problem costs them per month. If they cannot, the pilot has no scoreboard.
  • A single decision maker with budget. A plant manager with a discretionary budget beats a committee at headquarters.
  • Representative but not extreme conditions. You want a site that looks like your target market, not the hardest site they own.
  • Willingness to give access. You need to be on the floor, talk to operators, and pull data. A customer who will not allow that will not produce useful learning.
  • Reference potential. Ask early whether they would be named or take a reference call. A firm no should discount the value of the pilot.

Run two or three pilots in parallel if you can staff them, because a single pilot with a single customer teaches you that customer, not the market. Where the pilot sits in the larger sales motion is mapped out in the B2B sales process for a physical product.

Define Success in Writing Before Anything Ships

The single highest-leverage document in the entire engagement is one page long. It states, before the equipment arrives:

  • The specific metric and the threshold. Not "improve throughput" but "reduce changeover time from an average of 34 minutes to under 20 minutes across at least 40 changeovers."
  • How the measurement will be taken, by whom, and from what data source. Agree on the baseline before you start, because a customer's remembered baseline is always better than their actual one.
  • Start date, end date, and the number of units deployed.
  • What each side supplies: power, network access, operator time, training hours, an on-site contact.
  • What happens if the criteria are met. This is the clause founders omit and then regret.

That last item is the difference between a pilot and a demo. The document should say that on success, the customer intends to proceed to a purchase of a stated size, at stated pricing, within a stated window. It does not have to be a binding purchase commitment, but naming the number forces the conversation about budget while you still have leverage.

Charge for the Pilot

Free pilots get treated as free. A paid pilot, even at cost, changes the customer's internal posture: somebody had to justify the spend, which means somebody is now accountable for making it work. Typical structures are a fixed pilot fee covering hardware, installation, and support for the period, or hardware at cost plus a services fee.

A common and effective arrangement is to credit the pilot fee against the first production order. The customer gets the risk reduction, you get paid for the work, and the credit gives the conversion an obvious next step. Where the pilot pricing sits relative to your eventual list price matters too, since you are anchoring; think through your volume pricing tiers before you quote the pilot.

The Contract Terms That Actually Bite

Intellectual property. Corporate legal departments routinely send agreements assigning ownership of anything developed during the engagement to the customer. If your product improves because of pilot feedback, you must own the improvement. Push for background IP staying with each party and foreground IP staying with you, with the customer receiving a license to use the deliverables. The general landscape is covered in who owns the IP in product development.

Liability and warranty. Cap your liability at the pilot fee. Do not accept consequential damages for a pilot unit in an industrial process; a small company cannot carry that exposure. Say plainly, in writing, that pilot hardware is pre-production.

Exclusivity. Customers ask for it. Grant it only if it is narrow in field, short in time, and paid for with a minimum volume commitment.

Data rights. Secure the right to use operating data from the pilot to improve the product, in anonymized form, and the right to publish aggregate results.

Running It So It Converts

Instrument the units heavily. Remote logging means you learn about a fault before the customer calls, and gives you evidence when a result is disputed. Send a short written status every week, even when nothing happened, because silence reads as trouble. Respond within hours, not days; the buyer knows they are evaluating a supplier, not a device.

Hold a formal readout at the end with the data, the misses, and a specific proposal. Bring the champion the material they need to sell it internally: a one-page result summary, a cost justification, and an implementation plan. This is the same discipline that makes a hardware beta program produce usable results instead of anecdotes, and the reference you earn here is what opens the next ten conversations, including the booth traffic at your first trade show demo.

Get the Product Pilot-Ready

Projects House builds pilot-grade hardware that survives a customer site, with the logging, diagnostics, and serviceability a first deployment demands. Tell us about your product and the customer you are targeting through our contact form.