2nd/10_Wiki/Topics/Game_Design/Dynamic Offers.md

id, title, category, status, canonical_id, aliases, duplicate_of, source_trust_level, confidence_score, verification_status, tags, raw_sources, last_reinforced, github_commit, tech_stack

id	title	category	status	canonical_id	aliases	duplicate_of	source_trust_level	confidence_score	verification_status	tags	raw_sources	last_reinforced	github_commit	tech_stack
wiki-2026-0508-dynamic-offers	Dynamic Offers	10_Wiki/Topics	verified	self	dynamic-offer contextual-offer personalized-offer	none	A	0.9	applied	monetization mobile offers personalization live-ops		2026-05-10	pending	language framework typescript live-ops-offer

Dynamic Offers

매 한 줄

"매 player state 에 매 따라 매 individually-tailored 의 매 IAP bundle". 매 static catalog 의 evolution — 매 trigger event + 매 player feature + 매 ML scoring. 매 2026 mobile 의 매 default monetization surface — 매 LTV uplift 매 15-40%.

매 핵심

매 trigger 종류

• Lifecycle: 매 install D1, 매 first-defeat, 매 churn-risk.
• Progression: 매 level-up, 매 stage-clear, 매 stuck (3 fails).
• Event: 매 holiday, 매 weekend, 매 anniversary.
• Contextual: 매 low-resource, 매 PvP loss-streak.

매 offer 구성 요소

• Price tier: 매 $0.99 / $4.99 / $19.99 / $99.99.
• Contents: 매 currency + items + cosmetic.
• Value display: 매 "300% value" / "Best deal".
• Urgency: 매 timer, 매 limited inventory.
• Eligibility: 매 player feature gated.

매 응용

1. 매 starter pack (D1 high-conversion).
2. 매 stuck-pack (frustration relief).
3. 매 win-back (churned player 복귀).

💻 패턴

Offer eligibility engine

type OfferRule = {
  id: string;
  predicate: (p: PlayerState) => boolean;
  weight: number;
  cooldown_hours: number;
};

const RULES: OfferRule[] = [
  {
    id: "starter_pack",
    predicate: p => p.spend_total === 0 && p.session_count > 5,
    weight: 100,
    cooldown_hours: 24 * 7,
  },
  {
    id: "stuck_pack",
    predicate: p => p.consecutive_losses >= 3,
    weight: 50,
    cooldown_hours: 6,
  },
  {
    id: "winback",
    predicate: p => p.days_since_last_session >= 7 && p.spend_total > 0,
    weight: 80,
    cooldown_hours: 24 * 14,
  },
];

function selectOffer(p: PlayerState, now: number): OfferRule | null {
  const eligible = RULES
    .filter(r => r.predicate(p))
    .filter(r => !inCooldown(p, r.id, now, r.cooldown_hours));
  if (!eligible.length) return null;
  return weightedPick(eligible);
}

Price-tier model (player-specific)

function suggestPriceTier(p: PlayerState): number {
  const tiers = [0.99, 4.99, 9.99, 19.99, 49.99, 99.99];
  if (p.spend_total === 0) return 4.99; // starter
  const avg = p.spend_total / Math.max(1, p.purchase_count);
  // Pick tier near player's typical spend
  let best = tiers[0];
  for (const t of tiers) {
    if (Math.abs(t - avg) < Math.abs(best - avg)) best = t;
  }
  return best;
}

Bundle value calculator

type BundleItem = { id: string; quantity: number; reference_price: number };

function bundleValueRatio(items: BundleItem[], price: number): number {
  const total_ref = items.reduce((s, i) => s + i.quantity * i.reference_price, 0);
  return total_ref / price;
}

function valueLabel(ratio: number): string {
  if (ratio >= 5) return "BEST DEAL";
  if (ratio >= 3) return "GREAT VALUE";
  if (ratio >= 2) return "GOOD VALUE";
  return "";
}

Urgency timer rendering

class OfferTimer {
  expires_at: number;

  remainingLabel(now: number): string {
    const sec = Math.max(0, Math.floor((this.expires_at - now) / 1000));
    if (sec > 3600) return `${Math.floor(sec / 3600)}h ${Math.floor((sec % 3600) / 60)}m`;
    if (sec > 60) return `${Math.floor(sec / 60)}m ${sec % 60}s`;
    return `${sec}s`;
  }

  isExpired(now: number) {
    return now >= this.expires_at;
  }
}

Conversion-rate A/B harness

async function logOfferEvent(player_id: string, offer_id: string, event: "shown" | "clicked" | "purchased") {
  await analytics.track({
    user_id: player_id,
    event: `offer_${event}`,
    properties: { offer_id, ts: Date.now() },
  });
}

async function offerConversionRate(offer_id: string, since_ms: number) {
  const shown = await analytics.count({ event: "offer_shown", offer_id, since: since_ms });
  const purchased = await analytics.count({ event: "offer_purchased", offer_id, since: since_ms });
  return shown > 0 ? purchased / shown : 0;
}

매 결정 기준

상황	Approach
매 D0-1 player	Starter pack ($4.99, high value)
매 stuck player	Stuck-pack (resource focus)
매 churning	Winback (deep discount)
매 whale	Limited cosmetic + currency bundle

기본값: 매 trigger + eligibility + cooldown + 매 explainable rules first, ML overlay second.

🔗 Graph

• 부모: Game_Monetization_Strategy · LiveOps
• 응용: Capybara GO! · Data-Driven Personalization
• Adjacent: CPI (Cost Per Install)

🤖 LLM 활용

언제: 매 offer design, conversion analysis, rule engine setup. 언제 X: 매 cosmetic-only product — 매 dynamic 의 over-engineering.

❌ 안티패턴

• Predatory targeting: 매 frustrated whale 의 매 immediate offer.
• Cooldown 무시: 매 spam → 매 desensitization.
• Fake urgency: 매 reset timer → 매 trust loss.
• Value 거짓 라벨: 매 reference price inflation → 매 regulator risk.

🧪 검증 / 중복

• Verified (deltaDNA / Unity Gaming Services 2025 case studies).
• 신뢰도 A.

🕓 Changelog

날짜	변경
2026-05-08	Phase 1
2026-05-10	Manual cleanup — dynamic offers trigger + eligibility + value model.