CASE STUDYPAID MEDIA (META)

70x in eight months. Engineered, not accidental.

How Interconnections built a scaling operating system for Kalshi, the largest CFTC-regulated prediction market in the U.S., combining event-native creative velocity, structural consolidation, and truth-based measurement to scale Meta spend roughly 70x in 8 months.

CLIENT

Kalshi needed a scaling system, not just more budget.

Kalshi is a CFTC-regulated exchange in the U.S. that lets users trade on the outcomes of real-world events, including economic indicators, political developments, weather, sports, and cultural moments. Unlike traditional trading platforms, Kalshi demand is highly event-driven and attention-sensitive. Performance spikes correlate directly with real-world trends, which makes scaling fundamentally different from typical ecommerce or fintech. Interconnections was brought in to turn that demand into a predictable, scalable Meta operation.

  • IndustryPrediction Markets
  • MarketUnited States (CFTC-regulated)
  • Core ChannelMeta (Advantage+)
  • Primary KPICPCI (Checkout Initiated)
  1. 70×Spend Scaled
  2. 8 MonthsTime to Scale
  3. 8–9/dayNew Creatives
The Challenge

The problem was structural, not CPCI volatility.

At first glance, the issue appeared to be CPCI volatility and creative fatigue. Deeper analysis revealed structural problems that made predictable scaling impossible at the starting level, let alone at the scale Kalshi was aiming for.

  1. Attribution Misalignment

    Meta-reported Checkouts Initiated were inflated during key periods. Reported CPCI looked artificially low at times, so scaling decisions were being made on distorted data rather than on what was actually happening downstream.

  2. No Scaling Operating System

    The account lacked consolidation discipline, bid control guardrails, dayparting strategy, and intra-day monitoring protocol. At high budgets, these gaps become catastrophic.

  3. Creative Was Not Event-Native

    The platform is inherently event-driven. Generic creatives do not scale during attention spikes. The system needed daily creative velocity aligned with real-world trends.

  4. No Truth Validation Layer

    Without AppsFlyer cross-checks and backend reconciliation, the team had no way to verify Meta-reported numbers before committing major daily budgets at scale.

The Eight-Month Build

A chronological account of how the engine got built

Structured scaling began June 2025. Each month layered on a discipline: tracking truth, consolidation cadence, event-native creative velocity, dayparting, intra-day monitoring. The principle in play at each month is noted on the right.

  1. Jun '25

    Structured Scaling Begins

    Interconnections began structured scaling with unified objectives, truth-based measurement via AppsFlyer, and daily creative production. Established the operating system foundation.

    ~1.8× baselinePrinciple 03 · Truth-Based Measurement
  2. Jul '25

    Creative Velocity Ramps

    Two dedicated designers onboarded. Creative production hit 8 to 9 new assets per day. Event-native angles began outperforming generic creative significantly.

    ~2× baselinePrinciple 02 · Event-Native Creative Engine
  3. Aug '25

    Consolidation Discipline Pays Off

    Consolidation discipline proved its value. Fewer campaigns with stronger learning density. Dayparting implemented to cut overnight waste.

    ~2.4× baselinePrinciple 01 · Consolidation-First Structure
  4. Sep '25

    First Major Scale Jump

    NFL season kickoff drove massive event-based creative performance. Spend jumped roughly 3.4x in a single month. Bid controls kept CPCI within guardrails during aggressive scaling.

    ~8× baselinePrinciple 04 · Capital Control at Scale
  5. Oct '25

    Monitoring Becomes Mission-Critical

    Intra-day monitoring became critical at this scale. External UGC agencies brought on for additional creative capacity. Advantage+ placements confirmed as optimal after placement restriction tests failed.

    ~18× baselinePrinciple 04 · Capital Control at Scale
  6. Nov '25

    Election Cycle, Cultural Spikes

    Political event contracts drove attention spikes. However, political-heavy creative angles failed to sustain at this budget level. Sports and culture angles remained dominant.

    ~28× baselinePrinciple 02 · Event-Native Creative Engine
  7. Dec '25

    Record Checkout Volume

    CPCI optimized back down through consolidation and creative refresh cycles. Checkout volume roughly doubled from the prior month. The operating system proved its resilience at scale.

    ~46× baselinePrinciple 01 · Consolidation-First Structure
  8. Jan '26

    Peak Scale Reached

    Spend peaked at roughly 70x the pre-engagement baseline. The system Interconnections built was handling its largest daily budgets to date with disciplined control.

    ~70× baseline (peak)Principle 01–04 · All four principles compound
Operating Principles

The four principles that made 70x scaling work

No single tactic scales a Meta account 70x. Four principles ran in parallel: consolidation-first structure, an event-native creative engine, truth-based measurement, and capital control discipline.

  1. 01

    Consolidation-First Structure

    Fewer campaigns meant stronger learning density and tighter capital control. Consolidating every third day became one of the most important levers in the engagement.

    • Unified objective alignment to CI
    • Consolidation every 3rd day at scale
    • Reduced fragmentation, faster learning
    • Cleaner structure for intra-day moves
  2. 02

    Event-Native Creative Engine

    Creative velocity became the dominant growth driver. 8 to 9 new creatives per day across multiple formats, all tied to real-world event cycles.

    • Videos outperformed statics (~50% more spend)
    • Angles: NFL, NBA, NHL, weather, culture
    • UGC, head-to-head slips, news-style headlines
    • Daily batch launches, data-driven formats
  3. 03

    Truth-Based Measurement

    Corrected tracking before scaling. Meta was over-reporting Checkouts Initiated, which would have caused false scaling signals at high budgets.

    • AppsFlyer truth validation layer
    • Backend incremental conversion checks
    • Daily performance reconciliation
    • Multi-time-per-day reviews at peak scale
  4. 04

    Capital Control at Scale

    Scaling at this pace requires strict controls. Guardrails allowed rapid budget increases while protecting downside risk.

    • Dayparting: ads off 12AM–7AM
    • Cost caps and bid caps for CPCI guardrails
    • Advantage+ placements (restricted failed)
    • Intra-day monitoring for fatigue and drift
What We Did Not Do

The decisions that protected the scale

Scaling 70x is as much about what you refuse to do as what you execute. Four deliberate constraints kept the Interconnections system honest as budgets climbed.

  1. We did not scale on Meta-reported numbers

    Meta over-counted Checkouts Initiated, so platform CPCI was treated as directional only. Every scaling decision was validated against AppsFlyer and backend data first.

  2. We killed "get-rich-quick" creative

    High-CTR get-rich-quick angles pulled cheap clicks but weak downstream quality. Interconnections cut them after validation rather than scaling on vanity metrics.

  3. We dropped political angles at scale

    Political event creative worked in small pockets but failed to sustain at high budgets. Sports and culture angles carried the weight instead.

  4. We rejected restricted placements

    IG-only and FB-only placement tests underperformed. Advantage+ open delivery consistently won, so the system reverted and let Meta place the spend.

Results

70x spend growth, scaled with disciplined capital control

The trajectory tells the story. Structured scaling began in June 2025. Eight months later, monthly Meta spend had grown to roughly 70x the pre-engagement baseline, scaled with the capital controls the system was built to defend.

MonthSpend vs. BaselinePhase
June 20251.8×Foundation set
July 2025Creative velocity ramps
August 20252.4×Consolidation discipline
September 2025First major scale jump
October 202518×Intra-day monitoring
November 202528×Event and cultural spikes
December 202546×Record checkout volume
January 202670×Peak scale

Spend shown as a multiple of the pre-engagement baseline. Absolute spend, checkout, and cost-efficiency figures are withheld at the client’s discretion.

Key Takeaways

What this case proves

  1. 01

    Creative Velocity Is the Primary Scaling Lever

    8 to 9 new creatives per day, aligned with real-world events, ensured the algorithm always had fresh fuel. Without daily creative production, no amount of structural optimization would have sustained 70x scale.

  2. 02

    Truth-Based Measurement Prevents Catastrophic Errors

    Meta over-reported Checkouts Initiated during key periods. Without AppsFlyer validation and backend checks, scaling decisions would have been based on inflated data, which at this scale would have been catastrophic.

  3. 03

    Consolidation Beats Fragmentation at High Budgets

    Fewer campaigns meant stronger learning density, better algorithmic optimization, and tighter capital control. Consolidating every third day became one of the most important levers in the entire engagement.

RELATED PLAYBOOK

The system behind this scale.

This engagement runs on the Interconnections Meta ads scaling system, the same four-part operating system we deploy for every paid social scale.

FAQ

Common questions.

  • Interconnections built a scaling operating system for Kalshi on four principles: a consolidation-first campaign structure, an event-native creative engine producing 8 to 9 new creatives per day, truth-based measurement via AppsFlyer and backend reconciliation, and strict capital controls including dayparting and bid caps. Together these scaled monthly Meta spend roughly 70x in eight months.

Ready to build your scaling operating system?

This was not a tactic. It was an operating system. If your brand has demand potential but lacks scalable structure, Interconnections can help. Drop us an email. No pitch deck. No pressure. Just a real conversation about engineered growth.