How to get your first 10 activated users
A small startup does not need ten channels. It needs ten relevant people, a useful conversation, a clear first-value moment, and honest evidence about what happened next.
Define the result that counts
“Signed up” is usually not activation. Write one observable event that proves a single user received meaningful value. It should happen after registration and be difficult to trigger accidentally.
This proves access, not value.
This proves the reporting product completed its core job.
- Target user: the narrow group with the urgent problem
- First-value result: the concrete outcome one user completes
- Target window: one day, one week, or one month
- Evidence source: an authenticated backend event—not a manual guess
Choose two complementary learning channels
Use one channel where people already discuss the problem and one where you can research an individual before contacting them. This creates two different learning paths without spreading your attention across the internet.
Help where the problem is already visible
Look for current, specific questions on Reddit, Hacker News, Stack Overflow, Quora, or a focused community. Answer the question first. Mention your product only when it is genuinely relevant, and disclose your connection.
Research before writing
Choose people whose public work shows the problem may be relevant. Use a verified business route. Reference one real fact, ask one low-pressure question, and never guess an email address.
A useful content article can replace one channel when you already know a repeated question. A launch directory, ad, or viral loop is rarely the fastest first learning path unless you already have distribution.
Pre-register the two tests
Decide what you will do and when you will stop before seeing the results. This prevents a promising click or friendly reply from becoming a false success story.
| Decision | Write this before starting |
|---|---|
| Audience | One narrow role, situation, and problem |
| Offer | The useful help, pilot, or result you can honestly provide |
| Volume cap | Five relevant conversations or five researched people |
| Primary result | The confirmed activation event—not impressions or sends |
| Observation window | Enough time for a new user to reach first value |
| Stop conditions | Opt-out, complaint, hard bounce, community-rule conflict, or unsafe contact route |
Swipe horizontally to review every experiment-planning field.
Start conversations that are useful without a pitch
1. Answer the actual question with one concrete step.
2. Explain the tradeoff or failure mode.
3. Give a small example the reader can use now.
4. If relevant, disclose that you are building a related product and offer the link only when invited.
“Hi [name]—I saw [specific public fact]. I’m speaking with [narrow group] about [specific workflow]. How are you handling [one problem] today?”
Personalize only from source-backed facts. Do not add fake praise, urgency, outcome promises, or two questions disguised as one.One follow-up can be reasonable when the route and context permit it. Stop after a reply, opt-out, bounce, complaint, or the pre-registered sequence limit.
Help interested people reach first value
Early acquisition is inseparable from onboarding. Ask where the person got stuck, offer a consented setup session or precise guide, and fix one repeated obstacle. Do not identify individual users from analytics hashes or treat silence as permission to contact them.
- Give each experiment its own measured link.
- Preserve attribution from click through signup and activation.
- Separate recent users still inside the activation window from users who missed it.
- Test one onboarding change on a new cohort—never mix pre-change users into the result.
Review evidence, not activity
At the end of the window, compare the two small tests using denominators. Keep qualitative notes, but do not convert a compliment or manual status into a verified activation.
With only ten attempts, results are directional. Repeat a promising motion in another bounded cohort. Revise when people engage but do not activate. Stop a version that reaches its sample without a useful signal.
A practical 14-day schedule
- Day 1: confirm audience, offer, activation event, measurement, and safety boundaries.
- Days 2–3: find five current conversations and five source-backed prospects.
- Days 3–5: write useful replies and individual messages; review every destination and claim.
- Days 5–9: perform only approved actions and record delivered, published, failed, or unknown states honestly.
- Days 6–12: review replies, help interested people activate, and fix one repeated onboarding obstacle.
- Days 13–14: wait for the confirmed window, compare the two funnels, and choose one bounded repeat or a materially different test.
Give your product a supervised growth agent
UserGetter prepares the research, drafts, approvals, attribution, and evidence review. You stay in control of every external message, post, and dollar.
Build my first-user plan →