Meta Ads for UK raffle businesses
The Meta Ads playbook for a UK raffle business in 2026. Account structure, server-side CAPI, audiences, creative angles, and the CPA numbers operators are actually seeing.
TL;DR
- · Meta CAPI is mandatory in 2026. iOS 14+ killed browser-only pixels. Server-side conversion tracking is the only way Meta's algorithm can optimise effectively.
- · Run free-entry-route ads (not gambling ads). Mention the free-entry option in creative or on the landing page to stay ASA-compliant.
- · Realistic UK raffle CPA: £15-£40 in week 1, dropping to £8-£15 once Meta has 50+ pixel-confirmed purchases.
- · Start with broad-audience video creative, single-ad-set Advantage+ campaigns, and budget £30-£100/day to give the algorithm enough signal.
Before you start: prerequisites
Don't run a single ad until these five things are live. Skipping any of them wastes budget — your ads will run, but Meta's algorithm won't have the signal it needs to optimise.
- · Working checkout with Apple Pay + Google Pay. Mobile traffic is 70%+ — slow card-only checkout kills conversion.
- · Meta pixel + CAPI both deployed. Both. CAPI alone misses some events; pixel alone misses iOS users.
- · All key conversion events firing: ViewContent, AddToCart, InitiateCheckout, Purchase. Test in Meta Events Manager.
- · Free-entry route prominently displayed. ASA enforcement on hidden free entries has stopped operators in their tracks.
- · At least one live competition with stock + draw date. \"Coming soon\" landing pages won't run on Meta — needs real product.
Account structure that works
Meta's algorithm in 2026 favours consolidated account structures: fewer campaigns, broader audiences, more budget per ad set. The "20 ad sets each with £5/day" pattern from 2020 actively hurts your performance.
Recommended structure for a new raffle site: one ad account, one Business Manager, one Meta pixel, one CAPI dataset. Inside the account: 1-2 Advantage+ Shopping Campaigns optimising for Purchase, each with 2-3 ad sets (broad UK audience, lookalike if you have data, retargeting if you have warm pool).
Why this matters: Meta needs 50 purchase events per ad set per week to leave the learning phase. Spreading budget across 20 ad sets means none get enough signal; concentrating into 2-3 hits the learning threshold fast.
Naming convention: use a consistent pattern like "Prize-Name | Audience | Format" so you can sort by performance later. Example: "BMW M4 | Broad UK | Video".
CAPI: the non-negotiable
The Meta Conversions API (CAPI) sends conversion events server-to-server, bypassing iOS 14+ tracking limits and ad blockers. Without it, Meta sees 40-60% of your actual conversions. The algorithm optimises on bad data and your CPA spirals.
The Meta Conversions API official documentation walks through the implementation. The short version: when a customer completes checkout on your server, your server POSTs an event to Meta's API with hashed customer email/phone, the event details, and a unique event ID. Meta deduplicates with the pixel-side event and combines the signal.
Quality matters: Meta's Event Match Quality (EMQ) score grades how well your CAPI events match real users. Aim for 7+ out of 10. Send as much hashed customer data as you have (email, phone, name, postcode, IP, user agent, click ID).
Most modern raffle platforms ship with CAPI built in — Turbo IT's Launch tier sends server-side events for every conversion automatically.
Audience targeting
Meta's algorithm in 2026 is better than your manual targeting for cold acquisition. Go broad. Layer lookalikes once you have data. Save interest stacks for very specific niche prizes.
- · Cold acquisition (week 1-3): Broad UK 18-65, no interests, age and gender unrestricted unless prize is gender-specific. Let the algorithm find buyers.
- · Lookalike (week 4+): 1% UK lookalike of purchasers, top-tier. Needs 100+ purchase events seeded.
- · Retargeting (always-on): Last 30/60/90-day site visitors, AddToCart non-purchasers, video viewers. Cheaper CPA — these audiences already know you.
- · Avoid: Tiny interest stacks (e.g. "BMW + Top Gear fans + UK"), age over-restriction, geographic micro-targeting. All limit Meta's ability to find buyers.
Creative angles that convert
UK raffle ads that consistently win have three common features: prize is hero, social proof of past winners, and call to action is "enter from £1". Static images underperform video and motion in 2026 — the budget should skew 70%+ to video.
Winning angles:
- · Prize reveal video — 5-second hero shot of the prize, second 3 lifestyle context, second 8 ticket price + draw date. CTA: "Enter now".
- · Past winner UGC — real winner unboxing or reaction. Higher trust signal than studio creative.
- · Comparison framing — "Why pay £80,000 for this when you could win it from £1?" The friction is the price ceiling.
- · Urgency tied to draw date — "Last 48 hours" runs in the final 2 days of a competition. CPA halves in those windows.
Required compliance elements in creative: reference to free entry (or visible on landing page above-fold), 18+ messaging if prize is age-restricted, no misleading prize-value claims.
Realistic budget and CPA
| Phase | Daily budget | Typical CPA | Notes |
|---|---|---|---|
| Week 1 (learning) | £30-£100 | £25-£60 | Pixel cold, expect waste |
| Week 2-3 (optimising) | £100-£300 | £15-£35 | 50+ conversions seeded |
| Week 4+ (mature) | £300-£1,000+ | £8-£20 | Lookalikes + retargeting layered |
| Scale phase | £1,000-£10,000+ | £10-£25 plateaus | Diminishing returns set in |
The pattern is consistent: spend roughly £1,500-£3,000 in the first 30 days to get out of learning, then CPA stabilises around £10-£20 for the next several months until you hit your category's natural ceiling.
Scaling without breaking the algorithm
The thing that kills scaled raffle ad accounts: budget doubling. Going from £200/day to £500/day overnight resets the learning phase. Going from £200 to £250 doesn't. Scale by 20-30% per change, every 3-5 days.
- · Scale by 20-30% increments, every 3-5 days. Faster jumps trigger algorithm resets.
- · Refresh creative every 2-3 weeks. Same ad runs to the same audience too long = fatigue, frequency creeps up, CPA rises.
- · Don't kill ads in their learning phase just because CPA looks high. Give each ad set 7 days and 50+ events to stabilise.
- · Layer retargeting always-on — your warm pool is your highest-ROAS audience. Budget 10-20% of total spend on retargeting.
- · Add new prizes to refresh demand — running the same competition for 8 weeks is the single biggest ad-fatigue cause.
FAQ
Will Meta approve raffle ads?
Yes, when run as free-entry prize competitions (not gambling). Make sure the landing page prominently shows the free-entry option and includes 18+ messaging. Some ads get auto-flagged on first run — manual review usually clears them within 24 hours.
Do I need a gambling ad agreement?
No — that's only for licensed gambling operators. Free-entry prize competitions run as standard commerce ads.
How long until Meta starts spending profitably?
Roughly 2-4 weeks of consistent spend with proper CAPI. The pixel needs 50+ confirmed purchases before the algorithm has enough signal.
Can I run Meta ads without CAPI?
Technically yes. Practically your CPA will be 2-3x higher because Meta's optimising on incomplete data. CAPI isn't optional in 2026.
Should I run on Instagram or Facebook?
Both — let Meta auto-place across the family. Reels and Stories typically have lower CPMs but lower intent; Feed has higher intent but higher cost. The algorithm balances it.
Continue reading
Meta CAPI: server-side tracking
Why server-side conversion tracking is mandatory in 2026 — and how to implement it.
Instant-win games & the future of raffle sites
The conversion-lever your Meta ads will feed traffic into.
UK advertising rules for raffles
ASA-compliant copy patterns, free-entry disclosure, and 18+ messaging.
Pricing
Turbo IT ships CAPI server-side events for every checkout. £99/mo.
Ready to launch?
Launch your raffle site with Meta CAPI, Apple Pay, and the free-entry compliance plumbing all built-in. £99/mo + 19p per ticket, live in 1-2 weeks.