# A/B Test Plan — 1xBet Sri Lanka Prelanding

Minimum sample: **300 conversions per variant** before calling a winner.

## Test 1 — Hero Offer Angle

| Variant | Headline focus |
|---------|----------------|
| A (control) | 200% first deposit bonus + 150 free spins |
| B | Deposit from 500 LKR — play with welcome bonus |

**Primary metric:** CTA click rate, Registration CR  
**Secondary:** R2D (7d)

## Test 2 — CTA Copy

| Variant | CTA text (EN) |
|---------|---------------|
| A | Claim Bonus |
| B | Start Playing |

SI/TA equivalents must stay parallel.  
**Primary metric:** CTA CTR

## Test 3 — Hero Media

| Variant | Media |
|---------|-------|
| A | Static hero image (WebP) |
| B | Short muted loop video (only if LCP stays < 2.5s) |

**Primary metric:** Registration CR + LCP  
**Kill rule:** if B hurts LCP > 0.5s vs A, stop.

## Test 4 — Social Proof Style

| Variant | Block |
|---------|-------|
| A | Numbers ("12,000+ players in Sri Lanka") |
| B | 3 short player reviews |

**Primary metric:** Scroll depth 50% + CTA click

## Test 5 — Urgency Timer

| Variant | Treatment |
|---------|-----------|
| A | No timer |
| B | Countdown "Bonus ends in HH:MM:SS" (session-based, not fake forever) |

**Primary metric:** Registration CR  
**Compliance:** timer must not imply false scarcity if offer is ongoing — legal review.

## Implementation Notes

- Split via `?v=a|b` or `localStorage` key `ab_hero`
- Fire `ab_variant` event to GA4 on page load
- Keep UTM params identical across variants
- Run one primary test at a time on hero to avoid interaction effects