For my $5.50 lifetime ScsDriver WebDAV mount tool, the trial was leaking users. They signed up, poked around for 2 days, and left. The fix was a 3/3 feature cap + invite extension. Activation lifted +28%.
The 3 contradictions of a free trial
- Users want to use more → cost goes up
- Users don't know what to use → think product is "useless"
- Users hesitate at upgrade → churn
A 7-day trial makes all 3 worse. Users procrastinate, then cram on day 7, then leave without upgrading.
The 3/3 cap design
The trial is not "7 days, all features". The trial is "3 core features, capped at 3 uses each".
For ScsDriver, the 3 features are:
- Mount (connect to a WebDAV server)
- Sync (mirror local folder to remote)
- Automation (rules-based file moves)
Each feature can be used 3 times during the trial. After 3 uses, that feature is capped.
Mount: 1 / 3
Sync: 3 / 3 (capped)
Automation: 0 / 3
Why 3/3 beats 7 days
A 7-day trial has two failure modes:
- User remembers on day 7 → too late
- User does nothing days 1-6 → crams day 7 → performance issues
A 3/3 cap inverts both:
- User has to use features on day 1 (else they never hit 3)
- Each user is guided to all 3 features (not "find your own way")
Measured: 3/3 cap trial completion rate is 38% higher than 7-day trial.
The invite extension
When the user hits cap, give two choices:
- Upgrade ($5.50 one-time)
- Invite a friend (+7 days / +3 cap uses per friend)
You hit the cap (3 / 3)
[Upgrade lifetime $5.50] [Invite a friend, +7 days / +3 uses]
Don't hide the invite button.
Measured results:
- Upgrade rate: 22% (from the 4-step paywall funnel)
- Invite rate: 14% (invited users extend, half later upgrade)
- Net activation lift: +28%
Details that matter for invites
- Each friend adds +7 days OR +3 uses, not both. Pick one in the UI.
- Friend must actually register AND use 1 feature, not just sign up (anti-fraud)
- Inviter sees a progress bar: 3 / 5 invited → unlocks lifetime. Long-tail incentive.
- Invitee sees the inviter: "Your friend X invited you. You both get 7 days." Anti-spam framing.
Point 4 is critical. Don't make the invite feel like spam.
Failure modes I tried that didn't work
Cap 0/1 (each feature usable once):
- Trial completion: 12% (user can't learn)
- Invite rate: 4% (no motivation)
Cap 10/10 (essentially unlimited):
- Trial completion: 8% (overwhelming)
- Upgrade rate: 4% (user has no reason)
No invite button, upgrade-only:
- Upgrade rate: 9% (forced conversion)
- Negative review rate: 11% (user feels trapped)
The sweet spot is 3/3 + invite extension.
Why this compounds with the 4-step paywall funnel
The 4-step paywall funnel (visible counter + day-5 email + paywall + post-upgrade) handles "does the user know where they are". The 3/3 cap handles "does the user actually engage".
Both are needed. Without 3/3, the paywall fires on a user who's never used the product → they leave. Without the paywall, the 3/3 cap is just a limit with no next step.
Together: +57% conversion (paywall) × +28% activation (cap) = a meaningful product, not just a feature.
The takeaway
For tiny SaaS, the trial design is two layers:
- Layer 1 (cap 3/3 + invite): make sure users actually use the product
- Layer 2 (4-step paywall): make sure users know what to do next
Most products skip Layer 1. They have a paywall on a user who's never used the product. The user leaves.
For $5.50 lifetime ScsDriver, the 3/3 cap + invite extension is the single biggest activation lever I've shipped. It's also the lowest-effort, highest-impact change in the funnel.
If you only ship one change this year, ship the 3/3 cap.
Top comments (0)