Coniq
Back to the briefing

The Concrete Play · Coniq-specific

What Coniq should build.

Not a better generic loyalty engine — that lane is commoditising. The real-time intelligence & decisioning layer for physical destinations, monetising the cross-tenant data and presence signals no software-only competitor can reach.

01The Wedge

Build what only Coniq can.

Coniq isn't a generic loyalty SaaS — it's the platform for physical multi-venue destinations. That gives three assets a software-only competitor cannot assemble.

01

Location & presence intelligence

Why others can't · Software-only vendors have no geofence, app or sensor footprint — e-commerce loyalty has no concept of “is the shopper here right now.”

Real-time, location-triggered decisioning at the moment of physical presence.

02

Cross-tenant transaction data

Why others can't · DTC loyalty sees one brand; a CDP sees only what you pipe it. Neither has the multi-brand destination basket.

Destination-level intelligence and a retail-media network the operator can monetise.

03

The landlord-operator relationship

Why others can't · Pure SaaS sells to one brand’s marketing team — it has no B2B2B2C structure to plug media into.

A second revenue model (insights & media) that makes Coniq strategic, not just a tool.

The wedge: own the real-time destination-intelligence & decisioning layer — the things that need presence data, cross-tenant data and the operator relationship.

02What We Build

Four pillars, one wedge.

Each ties back to a strategic bet, each is grounded in an advantage only Coniq has, each is buildable on the rebuild stack. Tap a pillar.

The brain

Real-time decisioning at the edge of the venue

Bets #2 + #3 — but location-aware, which no e-commerce vendor can do.

A service that, given a member + live context (geofence dwell, POS/card-link transaction, app open, tier, balance), returns the ranked next-best-action in under 150ms — push, offer, points multiplier, challenge, tier nudge or partner offer.

  • Feature store fed by the live event spine
  • Contextual-bandit / RL core — reward = incremental visit/spend, constrained by margin & liability
  • Guardrails: consent/fatigue gate, per-offer margin caps, explainability log
  • Holdout & measurement harness baked in from day one

Build vs buy

BUILD the core — it’s the differentiator, and the obvious managed option (OfferFit) is now owned by Braze, a competitor. Buy the surrounding infra.

Proof metric

Incremental visit/spend lift vs holdout · decision latency p99 · offers/member within cap

03The Substrate

Cross-cutting enablers.

Not pillars — the foundation the pillars stand on. Specced in the rebuild blueprint, called out here for the product implications.

Transparent points ledger

Append-only & auditable — members and operators see why a balance changed. The anti-Starbucks-2026 trust posture.

Flexible mechanics engine

Paid tiers, experiential/status, gamification (streaks, missions), card-linked invisible earning, pooling, values-based redemption.

Breakage & liability intelligence

Predictive redemption + dynamic earn/burn as a first-class CFO tool. ~$10B/yr sits unspent — turn it into engagement.

Consent, identity & privacy

Send-time consent gate + crypto-shredding erasure. Compliance as architecture — and a prerequisite for retail-media.

04Sequencing

Layered onto the migration.

The rebuild is happening regardless. These build on it — ordered so each phase ships value before the next begins.

Now0–6 months
  1. 01Event spine + transparent ledger (penny-perfect migration)
  2. 02Native analytics — kill the Tableau dependency
  3. 03Agent-ready API contract
  4. 04One lighthouse real-time decisioning use case

Foundation that pays off immediately; native analytics is the fastest operator win.

Next6–18 months
  1. 01Generalised decisioning engine (bandits + margin/liability)
  2. 02Retail-media / destination intelligence product
  3. 03Operator copilot v1
  4. 04Flexible mechanics + breakage intelligence

The differentiation layer — the wedge becomes real and monetisable.

Later18 months +
  1. 01MCP server + UCP / ACP adapters
  2. 02Agent & voice/chat concierge surfaces
  3. 03Verifiable delegated identity
  4. 04Points interoperability across destinations

Optionality — cheap to add once the fabric exists, speculative to prioritise sooner.

05Discipline Is The Strategy

What we're not building.

Equally important. The anti-scope keeps us out of commodity fights and off the hype cycle.

A generic composable promo engine to out-feature Talon.One
Differentiate on location + cross-tenant data, not promo-primitive breadth.
Your own IdP, OLAP store, CDP, durable-execution or webhook system
Buy them. Engineering goes to the wedge.
Online agentic checkout as a near-term line for physical destinations
Build the primitives; skip the speculative surface.
Web3 / NFT loyalty
It corrected hard — only embedded utility survived.
A DTC / Shopify loyalty pivot
Different market, owned by others, abandons the destination wedge.
The Lighthouse

Prove the wedge on one operator first.

One anchor operator (Value Retail / MAG-class). Ship location-aware real-time next-best-offer + native analytics + one co-funded offer with 2–3 tenant retailers — measured on incremental visit/spend vs a holdout and the first retail-media dollar.

The Tesco Clubcard Challenges analog — but powered by physical presence. Prove it here, then generalise with Phase 2 de-risked.

The operator scorecard

How we'll know it's working

Speed

  • Decision latency p99 (<150ms)
  • Time-to-points
  • Integration time-to-live

Lift

  • Incremental spend vs holdout
  • Redemption rate
  • Second-visit rate

Economics

  • Retail-media revenue / operator
  • Breakage delta
  • Liability under management

Adoption

  • Operators self-serving
  • Copilot-assisted actions
  • Tenant retailers activated

Next chapter

Now — how do we build it?

Building in the agent era: the operating model behind the wedge.

How we build