It's emotion & utility (why a human comes back), AI decisioning (the right value to each person, in real time), and machine-readability — so an AI agent can act on that value at the moment of purchase.
The platforms that win the next decade are AI-native, real-time, composable and agent-ready — the exact bar today's batch-architected incumbents cannot meet. That is the warning, and the opening, for Coniq.
Global retail revenue AI agents could orchestrate by 2030
Source · McKinsey
US loyalty points left unspent every year — pure lost engagement
Source · Antavo
Offer-decision speed the agentic era demands. Batch systems can't.
Source · Eagle Eye
Not one trend — three simultaneous shifts in what loyalty is, how it decides, and where it happens.
Paid membership, status, experiential access and emotion now out-perform discounting. A good points program is table stakes, not a differentiator.
Proof · Paid-program members are 60% more likely to spend more (vs 30% for free). Special access is now the #1 driver of program value.
Contextual bandits and reinforcement learning pick the best offer for each individual, continuously. Loyalty data is the post-cookie moat; AI is the engine on top.
Proof · 71% of marketers are raising loyalty investment specifically because third-party cookies are dying. Owning consented data + decisioning is the AI-native edge.
Agents now do discovery and checkout on the customer's behalf. If your loyalty program isn't machine-readable, your value is invisible at the moment of purchase.
Proof · Checkout rails shipped fast (OpenAI ACP, Google UCP, Visa & Mastercard). Loyalty is 6–12 months behind — the risk and the opening.
Loyalty is now over half the marketing budget and returns 5.3x — yet most members feel nothing, and billions in value evaporate unspent.
programs joined
…but only ~8.8 actively used. Membership ≠ engagement.
Bond
more likely to spend
Paid-program members vs 30% for free programs.
McKinsey
higher CLV
Subscription-loyalty members vs non-members.
McKinsey
average loyalty ROI
83% of programs report positive ROI in 2026.
Antavo
of marketing budget
Now spent on loyalty & CRM — up from 22.8% in 2022.
Antavo
unspent points / yr
Over 25% of US points never redeemed — a liability and lost engagement.
Antavo
The emotional gap
A 26-point emotional gap — the headline opportunity. · Antavo GCLR 2026
Market trajectory
~$13B (2025) → $20–31B by 2030, depending on analyst. The adjacent CDP market grows even faster (30.7% CAGR) — and is converging with loyalty.
When an agent does discovery and checkout, the brand may never own the moment of consideration. The plumbing shipped fast in 2025–26. Loyalty is 6–12 months behind.
What actually shipped
Instant Checkout & the Agentic Commerce Protocol (ACP) ship inside ChatGPT.
AP2 — Agent Payments Protocol — cryptographically signed purchase mandates.
Agent Pay: 'Agentic Tokens' bind a credential to cardholder + agent + consent.
Rufus gains autonomous auto-buy; folds into 'Alexa for Shopping' (May 2026).
Universal Commerce Protocol (UCP) — agent shopping with Target, Walmart, Shopify.
Unified Incentives Protocol — first standard to make loyalty visible to AI agents.
OAuth identity-linking carries member pricing into an agent cart — perks preserved.
First fashion brand to check out inside Gemini — but loyalty was explicitly left out.
Share of US e-commerce via agents · 2030
McKinsey: agents could orchestrate ~$1T US / $3–5T global retail revenue. AI traffic to retail sites grew +4,700% YoY, then +393%.
“Agentic commerce is the new race track — but most loyalty programs are still in the garage.”
The canonical tell: Gap × Gemini (Mar 2026) became the first fashion brand to check out inside an AI assistant — and loyalty was explicitly left out.Checkout shipped; loyalty deferred. A benefit an agent can't see is invisible at purchase.
Contextual bandits and reinforcement learning pick the best offer for each person, continuously — replacing segment-level A/B testing. This is shipping at scale today.
Acquired for $325M → BrazeAI Decisioning Studio. A 'community of bandits' optimises channel, timing, creative & offer per person.
Result
Kayo Sports scaled 300 → 1.2M daily message variations, +14% reactivation, +105% cross-sell.
Eagle Eye said it of its own market in March 2026: 'Most loyalty programs aren't ready for AI agents.' Batch-architected, segment-based incumbents structurally can't meet the new bar. That gap is the opportunity.
Consolidation is squeezing the middle — Zeta absorbed Marigold, Capillary IPO'd, Talon.One raised $135M. The open lanes are AI-native, real-time and composable.
Eight bets that take us beyond rebuilding what we have, toward the AI-native, real-time, agent-ready platform the next decade rewards. Tap any bet to see why it matters.
The through-line
Win by combining emotion & utility, AI decisioning, and machine-readability — so the right value reaches each person, and an agent can act on it at the moment of purchase.