Growth experiments — SaaS playbook (e5)
Practical experiments and tested UX patterns to lift activation, onboarding conversion, and early retention for software products. Lightweight, measurable and tailored for B2B SaaS.
- Focus: Activation, First-value, Time-to-Aha
- Approach: Rapid hypotheses, cohort measurement, UI/UX variations
- Delivery: Design assets, experiment specs, success criteria
Experiment framework
1. Hypothesis
Clear statement tying change to expected metric movement.
2. Design variant
Mockups, user flows and implementation notes for engineers.
3. Measurement
Primary KPI, sample size, confidence, and guardrails.
When to run experiments
Run quick UI/UX changes early for activation and medium-duration flows for retention. Use feature flags to reduce rollout risk.
Top experiments (accordion)
Click each experiment to reveal hypothesis, design notes and expected KPI.
Case study: Onboarding checklist (SMB SaaS)
Overview
We launched an in-app onboarding checklist that highlighted 3 core actions. Experiment ran for 6 weeks on new sign-ups.
- Variant: Checklist with progress and skip control
- Result: +18% activation, +11% day-14 retention
Metrics & playbook
Key metrics we track per experiment and quick playbook items for each outcome.
Primary metrics
- Activation conversion (Day 0–3)
- Time-to-Aha (median minutes)
- Retention (Day 7 / Day 14)
Playbook
- Stop or iterate if *** lift at 95% CI
- Scale rollout with feature flags and monitoring
- Document learnings and update onboarding flows
Team & lead
Ava Morgan
Ava leads cross-functional experiments and designs measurable UX interventions for product-led teams.
ConnectHow we work with product teams
We embed with PMs and engineers for short sprints, deliver Figma-ready variants, and provide clear success criteria so A/B results are actionable.
FAQ & quick links
Common questions
Q: How long to see results?
A: Small UI experiments often show signals within 2–3 weeks; larger flows need 4–8 weeks.
Q: Do you implement experiments?
A: We deliver designs, specs and can support implementation with partner engineers.
Q: Are results shareable?
A: Yes — we provide exportable reports and playbook updates.