Where do the first real users come from?
Start with reachable people facing the problem, and make a trial produce observable feedback.
On this page
Direct answer
Find people most likely to face the target problem, help them complete the task, then expand exposure. Start with personal use, relevant groups, useful public development notes and suitable launch channels, feeding observations back into the product. A cold start should discover both users and reasons the product does not yet work. Visits without task completion or follow-up do little to guide a next step. The channel ordering and records below are practical suggestions.
When to use
- A new product has few visitors or registrations.
- Some traffic exists, but you do not know whether visitors are target users.
Step by step
List places where the users already gather
Start with interviewees, existing business relationships, professional groups, discussions and search questions. For each channel, explain audience fit, permitted introduction and concrete help; prefer channels where feedback is observable. Build a reach table: channel, recent relevant discussion, who is doing the task, contact rules and why you help. Narrow a generic peer group to actual buyers, such as shop owners preparing listing images. Peer feedback is not buyer payment. Existing interviewees and people actively seeking help are enough to start; a large following is unnecessary. Look for real error reports for developer tools, listing and operations discussions for merchant tools, résumé/application tasks for career tools, or niche software discussions for personal tools. Peers on launch platforms are not automatically buyers. Find recent requests first, then check current self-promotion rules; do not choose a platform solely by audience size.
Prepare an experience that completes a task
Explain in one sentence who gets which result. Prepare representative input, an actual output demonstration and clear limits. Reduce unnecessary signup/configuration without removing identity, privacy or cost protections. Before publishing, let someone unfamiliar with the product try its core path.
Make a concrete, honest trial invitation
Respond to a problem the person has mentioned and disclose your role and actual development stage. Ask for feedback on a real task rather than vague support. Respect community rules, contact preferences and refusal; do not send bulk unwanted invitations. Check whether the person recently performed the task, expects it again and helps decide purchases. Make one specific request, such as observing a real workflow or trying a sample; state the time required and freedom to decline. No response is not proof of no demand and does not justify endless follow-ups.
Observe use and follow up on obstacles
Record starts, completion, departure points and whether the result was used. Ask about the task rather than only aesthetic preference. Tell people who reported problems what changed, building a continuing relationship.
Turn a working path into the next experiment
Compare qualified users, task completion and payment across channels. Share informative process, failures and improvements, not just code-volume updates. Expand launch only after the core path is stable. Offer a voluntary way to keep in touch and explain what will be sent; do not automatically add trial participants to marketing. Separate preparation, first invitation, delivery follow-up and review. Check channel/demo first, record contact counts, ask about adoption at the agreed time, then compare the same cohort. No replies suggests list, context or invitation problems; starts without completion call for delivery repair; adoption without buying requires checking buyer, value, price and frequency. Choose one principal problem per round.
Worked example
Hypothetical example: a tool for independent merchants
Suppose you find a shop owner preparing new listings in a familiar merchant group and, within its rules, invite a trial with real images. You discover a format mismatch, fix it and show the improvement. A truthful before-and-after demonstration with limits can attract another person with the same task more effectively than sending the same advertisement to unrelated groups. This scenario is not evidence of an actual outreach campaign.
Action checklist
Reusable templates
Cold-start experiment plan
Target users and current problem:
First channel and fit:
Rules and permitted contact:
Complete task available to try:
Demo, limits and support:
Observe start / completion / adoption / payment:
Feedback record:
Review date:
Evidence needed to continue this channel:
Trial invitation and follow-up wording
[First invitation: fill in real context]
Hello, I saw you mention [problem] in [specific discussion/situation]. I am building [result for people], currently at [actual stage].
If [task] is coming up, would you like to try it with [necessary, anonymizable material]? It would take [time] from you; I would deliver [result] by [deadline]. Current limits are [limits], and the cost is [free/explicit offer]. It is fine to decline.
[Follow up at an agreed time]
Did you use the result? Which step still needed your work? When will the task recur? If you did not adopt it, why?
Do not add trial participants to marketing without consent; stop contacting someone who declines.
First-user channel ledger
Period and target task:
Channel | relevant discussion | audience | contact rules | invitation version | reached | replied | eligible | started | adopted | paid | follow-up date
Deduplication of the same person across channels:
Possible reasons for no response:
Reasons for replying but not starting:
Reasons for completing but not adopting/paying:
Continue/pause channels and evidence:
One change next round:
Channel-fit exercise: find discussion before inviting
Target role and recent task:
Two possible gathering places:
Evidence: discussion link / time / specific difficulty / alternative / possible buying role
People who do not fit this product:
Self-promotion, direct-message and disclosure rules, with review date:
One sample result I can provide:
Available capacity and response time:
Stop: refusal / rules prohibit / unreliable delivery / exhausted budget
What effort or adoption would turn this channel lead into demand evidence?
Pitfalls
- Treating launch-day attention as durable growth.
- Asking peers for likes and treating that as target-user feedback.
- Expanding traffic before reliable delivery, magnifying failures and cost.