The Short Answer
What is a good MVP launch strategy?
Launch to a narrow group of people who have the problem badly, recruit the first ones personally, and release in stages rather than all at once. Track the one success metric from the first day, talk to every early user, and judge the launch on whether it answered your hypothesis, not on sign-up numbers.
What an MVP Launch Is For
A big launch day feels like progress, but for an MVP it often works against you: lots of curious visitors who do not have the problem, little time to talk to any of them, and data that says more about the launch than about the product. The aim is a steady flow of the right users and a clear answer to the question in your scope document.
| Marketing launch | MVP launch | |
|---|---|---|
| Goal | Attention and sign-ups | An answer to the hypothesis |
| Audience | As wide as possible | People with the problem, as narrow as possible |
| Pace | One big day | Staged, small groups |
| Success measure | Sign-ups, visits | The success metric set before launch |
| Founder’s job | Promotion | Talking to every early user |
Who to Launch to First
The best first users have the problem often, already spend money or time working around it, and can be reached personally. In discovery you should have met some of them; start there. A good test: could you name the first 20 people you will invite? If not, the audience is still too broad.
Avoid launching only to friends and family. They will use the product to be kind and say positive things for the same reason, and neither tells you whether strangers want it.
Recruit the First Users by Hand
How do you get the first users for an MVP?
Personally. Message, call or visit the people you met in discovery, offer to set them up, and help them get value the first time. Paul Graham calls recruiting users manually the most common unscalable thing founders have to do at the start.
In his essay Do Things That Don't Scale (July 2013), Graham describes Airbnb's founders going door to door in New York to recruit hosts and improve their listings, and Stripe's founders setting new users up on the spot rather than sending a link. The point is not the tactic; it is that early on, the founders are the distribution.
Manual onboarding has a second benefit: watching someone use the product for the first time is the fastest usability test you will ever run.
Launch in Stages
Release to a small group, fix what breaks, then widen. GOV.UK runs the same pattern for public services: a private beta with a limited number of invited users, then a public beta once the team is confident it can run the service at scale. Mondo did the same in 2015, starting with a first wave of 500 alpha testers before widening.
- 1
Founder-led group (10–20 users)
People you recruited personally. Onboard each one yourself and watch where they struggle.
- 2
Invite-only group (50–100 users)
Referrals and a waiting list. The product should now work without you in the room.
- 3
Open launch
Public listing, communities and launch sites, once the core flow is reliable and you know your conversion rates.
For mobile apps, allow for store rules: Apple reviews 90% of submissions in under 48 hours, and new personal Google Play accounts must run a 14-day closed test with at least 12 testers before publishing. For running an invite-only group well, see running a beta program.
Size Your Launch
Work backwards from the number of active users you need to answer the question. The product-market fit survey becomes directionally useful at about 40 respondents, so many MVPs need somewhere around 100 active users.
Interactive calculator
How many people do you need to reach?
Work backwards from the number of active users you want. Use your own conversion rates; the defaults are placeholders, not benchmarks.
5,000
Visitors to reach
250
Sign-ups needed
100
Active users
Sign-ups = active users ÷ activation rate. Visitors = sign-ups ÷ sign-up rate. Replace the defaults with your own numbers after the first week of launch; your real rates are the most useful data the launch produces.
What to Measure From Day One
| Metric | What it tells you |
|---|---|
| Activation: users who reach the core value once | Whether the product is understandable and the onboarding works |
| Repeat use in week two | Whether the value is real, not novelty |
| Conversion to paid, if you charge | Willingness to pay, the strongest signal there is |
| Where users drop out of the core flow | What to fix first |
| The success metric you set before launch | The answer to the hypothesis |
Analytics on the core flow should be a Must Have in the build, not something added after launch. Once you have enough active users, run the survey in what to do after the MVP.
A Four-Week Launch Plan
| Week | Focus | Done when |
|---|---|---|
| 1 | Founder-led onboarding of the first 10–20 users | Each has reached the core value at least once |
| 2 | Fix the biggest drop-off; invite the next group | Activation is stable for new users without help |
| 3 | Invite-only growth through referrals and the waiting list | You know your sign-up and activation rates |
| 4 | Survey active users; compare results with the target | A written persevere, pivot or stop decision |
This is an example plan, not a benchmark; adjust the group sizes to the number of users your metric needs. For the build before launch, see the MVP development process, and for a fixed-price build ready to launch in 3 to 6 weeks, see MVP development.
Frequently Asked Questions
How do you launch an MVP?
To a narrow group of people who have the problem, recruited personally, in stages: a founder-led group, then an invite-only group, then an open launch. Track the success metric from day one and talk to every early user.
How do you get your first users for an MVP?
By hand. Contact the people you met in discovery, offer to set them up and help them get value the first time. Paul Graham calls recruiting users manually the most common unscalable thing founders have to do at the start.
Should I launch my MVP on Product Hunt?
Later rather than first. Launch sites bring many visitors who do not have the problem. They work best once the core flow is reliable and you know your conversion rates, so the extra attention turns into useful users.
How many users do you need to validate an MVP?
Enough to measure your success metric and run a product-market fit survey; Superhuman found survey results became directionally correct at about 40 respondents. Many MVPs therefore aim for around 100 active users.
What should you measure after launching an MVP?
Activation, repeat use in week two, conversion to paid if you charge, where users drop out of the core flow, and the success metric you set before launch.
Should an MVP launch be public or private?
Start private. A small invited group lets you fix problems and talk to users before a wider release. GOV.UK runs public services the same way, with a private beta before a public one.
A good MVP launch is quiet, personal and measured. The loud launch can wait until you know what you are launching.