/pulse

Activation Rate Design Reference

Purpose: Aha-moment discovery, Magic Number identification, time-to-value (TTV) measurement, and activation milestone design. Produces an evidence-grounded activation contract that precedes retention in the lifecycle funnel.

Scope Boundary

• pulse activation: Aha-moment / Magic Number discovery, activation event design, TTV, activated-vs-not retention overlay (this document).
• pulse retention (elsewhere): Retention curves and Power User analysis. Activation feeds retention; they are measured as a linked pair.
• pulse funnel (elsewhere): Generic conversion funnel. Activation is one specific conversion step inside that funnel.
• Onboarding copy (Prose onboarding): UX writing for the activation flow. Activation measurement owned by Pulse; activation copy owned by Prose.
• Growth (elsewhere): Post-activation habit formation and re-engagement. Activation is pre-Day-7; Growth is Day-7 onwards.

Workflow

DISCOVER  →  compare activated vs non-activated cohorts
          →  surface behaviors that predict retention ≥ 1.5x baseline
          →  candidate events from the "activation suspect list"

MAGIC     →  find the quantitative threshold (N actions, M days)
          →  e.g., "7 friends in 10 days", "2k messages", "5 invoices/mo"
          →  validate via retention lift and regression

DEFINE    →  commit the activation event (name, payload, window)
          →  set target rate (B2C 20-40%, SaaS self-serve 50-70%, PLG top 75%+)
          →  set TTV target (SaaS <7 days, consumer <1 session)

FUNNEL    →  signup → first-value action → activation milestone
          →  measure drop-off per step; instrument events

OVERLAY   →  activated cohort retention vs non-activated (W1/W4/W12)
          →  segment by acquisition channel × plan tier

CONTRACT  →  publish activation registry; alert on drift
          →  handoff to onboarding design (Prose) and experiment (Experiment)

Aha-Moment vs Magic Number

• Aha-moment: qualitative experience where the user perceives core value. Discovered via user research (Field), session replay (Trace), and first-session analytics.
• Magic Number: quantitative threshold that operationalizes the aha-moment. A measurable threshold the product team can move.

The Magic Number must be:

1. Behavioral (user performs action), not demographic (user is in segment X).
2. Predictive (activated cohort retains ≥1.5x baseline at Week 4+).
3. Achievable within TTV window (most new users can reach it in < target TTV).
4. Specific and testable (not "uses the app regularly" — "sends ≥3 messages within 48h").

Industry Examples

Activation Rate Benchmarks (2024-2026)

AI-product caveat: an activation event for an agent product should be outcome-grounded ("first task completed with user-confirmed success"), not "first prompt sent" (which is the AI-native equivalent of signup_completed — fires on intent, not on value). Pair with an offline eval score to detect activation rates inflated by users who "completed" a low-quality output.

Event Schema

interface ActivationEvent {
  event_name: "activation_reached";
  user_id: string;
  activation_type: "magic_number" | "milestone" | "aha_moment";
  magic_number_threshold: number;  // e.g., 7 for "7 friends"
  magic_number_actual: number;     // achieved count
  time_to_activation_sec: number;  // TTV from signup
  acquisition_channel: string;
  plan_tier: string;
  first_value_action_at: string;   // ISO timestamp of first value event
  activated_at: string;            // ISO timestamp of activation
}

Anti-pattern: naming this event "user_activated" without the numeric evidence makes it impossible to audit or re-tune when the Magic Number shifts.

Funnel Instrumentation

SIGNUP
   │
   ├─[signup_completed] ─ t=0
   │
   ▼
ACCOUNT SETUP
   │
   ├─[profile_completed]
   ├─[first_workspace_created]
   │
   ▼
FIRST VALUE ACTION (milestone 1)
   │
   ├─[first_value_action]              ← must be logged separately
   │                                     (e.g., first_message_sent)
   ▼
MAGIC NUMBER PROGRESS
   │
   ├─[magic_progress_1] ─ threshold * 0.25
   ├─[magic_progress_2] ─ threshold * 0.50
   ├─[magic_progress_3] ─ threshold * 0.75
   │
   ▼
ACTIVATION
   │
   └─[activation_reached]              ← retention predictive threshold

SQL — Activated vs Non-Activated Retention

WITH activated_cohort AS (
  SELECT
    user_id,
    DATE(activated_at) AS activation_date
  FROM events
  WHERE event_name = 'activation_reached'
),
all_signups AS (
  SELECT
    user_id,
    DATE(signed_up_at) AS signup_date
  FROM users
),
cohort_labels AS (
  SELECT
    s.user_id,
    s.signup_date,
    CASE WHEN a.user_id IS NOT NULL THEN 'activated' ELSE 'not_activated' END AS status
  FROM all_signups s
  LEFT JOIN activated_cohort a USING (user_id)
)
SELECT
  c.status,
  DATE_DIFF('week', c.signup_date, DATE(e.event_at)) AS weeks_since_signup,
  COUNT(DISTINCT e.user_id) * 1.0 /
    COUNT(DISTINCT c.user_id) OVER (PARTITION BY c.status) AS retention_rate
FROM cohort_labels c
LEFT JOIN events e
  ON e.user_id = c.user_id
  AND e.event_name = 'nsm_action_completed'
  AND e.event_at >= c.signup_date
GROUP BY c.status, weeks_since_signup
ORDER BY weeks_since_signup;

Expected signal: activated cohort retains ≥1.5x (preferably ≥2x) vs non-activated at Week 4+. If lift < 1.5x, the Magic Number is not yet predictive — refine the threshold or event definition.

Segment Cuts

Always report activation rate broken down by:

1. Acquisition channel (organic / paid / referral / direct) — paid traffic often activates lower.
2. Plan tier (free / trial / paid) — paid usually activates higher; intentional, not a warning.
3. Signup cohort week — watch for week-over-week drift (≥5pp drop = alert).
4. Persona / role (if self-reported) — B2B SaaS varies 2x-5x across persona.
5. Device / platform (mobile / desktop) — mobile often activates faster but shallower.

Activation Drift Alerts

Activation Registry Template

## Activation: [Product]
- **Aha-moment (qualitative)**: [user perception]
- **Magic Number (quantitative)**: [N actions] within [M time window]
- **Event name**: activation_reached
- **Target rate**: [e.g., 60% in 7 days]
- **Current rate**: [%]
- **TTV target / actual**: [e.g., <7d / median 4.2d]
- **Retention lift (W4)**: [e.g., 2.3x vs non-activated]
- **Owner**: [team]
- **Last reviewed**: [date]
- **Known gaming vectors**: [e.g., automated signup + bulk action]

Deliverable Contract

When activation completes, emit:

• Aha-moment statement (qualitative one-liner from research).
• Magic Number (quantitative threshold, window, evidence of retention lift).
• Activation event schema (typed, with payload contract).
• Funnel instrumentation plan (signup → first value → progress → activation).
• Activated-vs-not retention comparison (W1/W4/W12 with lift ratio).
• Segment cuts (channel, plan, device).
• Drift alerts with thresholds and severity.
• Activation Registry entry (template above).
• Handoff targets: Prose for onboarding copy, Experiment for uplift testing, Growth for post-activation engagement, Field for qualitative validation.

References

• Sean Ellis — "Hacking Growth" (activation definition + north star)
• Hila Qu — "The Art of PLG Activation" and Reforge PLG course
• Facebook / Chamath Palihapitiya — "7 friends in 10 days" (Stanford talk, 2015)
• Andrew Chen — "New Data Shows Losing 80% of Mobile Users is Normal"
• Amplitude — Product-Led Growth Playbook (activation benchmarks)
• Appcues — PLG Benchmarks Report 2024 (TTV, activation rate)