In short: in a market that does not exist yet, first customers are chosen as design partners and references at once, bought through institutional procurement, reached through practitioner networks, and measured by risk retired.
Key points
- Design partners should own the cost, resemble the wider market, and agree in pilot terms to be a reference.
- Write success criteria, a sponsor, a budget and an end date into every pilot agreement.
- Prepare procurement routes, a security pack, insurance and planning calendars before approaching institutions.
- Treat consortia, standards bodies and regulator sandboxes as distribution, not PR.
- Report risk retired monthly: field milestones, paid pilots, security reviews, clearances and written references.
Design partners who are also references
Part 7 said the first ten customers are design partners. In long cycles that rule tightens. Each design partner costs a year or more of attention. You cannot afford one who shapes the product and then will not say so.
Choose design partners on three tests. They own the cost from Part 31. They can deploy in a setting that looks like the wider market. And they agree up front, in the pilot terms, to act as a reference if the agreed success criteria are met.
Write the success criteria into the pilot agreement. Vague pilots never end. Specific pilots convert or close, and both are useful.
Expect fewer partners than a software company would run. Three good ones, each in a different buyer type, teach more than ten shallow ones.
Pilots with governments, regulators, utilities and large corporates
These are the natural first buyers for robotics, infrastructure, energy, identity and verification. They are also the slowest. Four things differ from selling to a startup.
| Area | What differs | What to prepare |
|---|---|---|
| Procurement | Formal process, frameworks, thresholds, often tender rules | Map the route before first contact; find the threshold that allows a direct pilot |
| Security | Review before any data or site access | Security pack, data handling, access controls, ready on day one |
| Liability | Risk sits with the vendor; insurance and indemnity are negotiated | Insurance in place; a lawyer who has seen the buyer's standard terms |
| Timelines | Budget cycles, annual plans, committee calendars | Know their planning calendar and work back from it |
The lesson is to sell the pilot as a project the institution can approve, not as a trial it can tolerate. That means a budget, a sponsor, a procurement route and an end date. Pilots without those are favours, and favours do not convert.
Part 4 said run the round like a sales process. Run institutional pilots the same way: stages, owners, exit criteria.
Consortia and standards bodies as distribution
Part 8 said find channels that are open when platforms are closed. In emerging tech the open channel is often an industry consortium, a standards working group or a regulator's sandbox.
They are slow and unglamorous. They are also where the buyers sit when they are deciding what the category means. A company in the working group that writes the standard is a company every member has met.
Treat them as distribution, not PR. Pick the one where your buyers are, not the one with the best logo. Send the person who can speak to the technical detail. Show up for the whole cycle, not the launch.
The same applies to regulator sandboxes. A sandbox place is a pilot with a clearance attached, which Part 32 said is a milestone.
The first hundred are practitioners, not a public community
Part 6 said the first hundred people are the ones who will tell others. Part 5 said community is a job. In emerging tech the hundred are not a public community at all.
They are the field engineers, grid operators, compliance leads, security researchers and procurement specialists who decide whether a new thing is credible. They do not gather on social platforms. They gather in working groups, technical forums, conferences with small audiences and private channels.
Build the list by name. Know what each person needs to see before they will vouch for you. Give it to them before you ask for anything.
Part 9 said stop reporting reach. Here that means a hundred practitioners who will return a call matters more than ten thousand followers who will not.
Measure progress by risk retired
Pipeline value is the wrong measure when deals take two years. A large pipeline of unconverted pilots is a cost, not an asset.
Measure what reduces risk. Technical milestones met in the field. Paid pilots signed and paid. Security reviews passed. Regulatory clearances won. References agreed in writing. Each one makes the next sale faster and the next raise cheaper.
Report those monthly, to the team and the board. Part 10's ten-question self-score works here with one change: replace revenue questions with risk-retired questions until revenue arrives.
The checklist
- [ ] Each design partner owns the cost, resembles the wider market and has agreed in writing to be a reference
- [ ] Pilot agreements name success criteria, a sponsor, a budget and an end date
- [ ] Procurement route, security pack, insurance and timeline are prepared per buyer type
- [ ] One consortium, standards body or sandbox is chosen where the buyers sit
- [ ] The first hundred practitioners are listed by name with what each needs to see
- [ ] Monthly reporting tracks risk retired and paid pilots, not pipeline size
Common questions
How do I find first customers for a product in a brand-new market?
Choose a few design partners who own the cost, can deploy in a setting like the wider market, and agree up front to be references if agreed success criteria are met.
How should a startup run a pilot with a government or large corporate?
Sell it as a project the institution can approve, with a budget, a sponsor, a procurement route and an end date. Prepare security, liability and timeline answers before first contact.
How should emerging tech startups measure progress when sales cycles are long?
Measure risk retired rather than pipeline value: field milestones met, paid pilots signed and paid, security reviews passed, regulatory clearances won and references agreed in writing.
Are standards bodies worth a startup's time?
Yes, as distribution. They are slow, but buyers sit there while deciding what the category means. Pick the one where your buyers are and attend the whole cycle.