Demo-First Verification

Demo-first verification operationalises the principle that vendor verification is necessary but not sufficient for buyer-side trust. The vendor's published live track record reflects the EA's behaviour on the vendor's broker under the vendor's deployment conditions. The buyer's realised performance will differ to the degree that broker conditions, VPS infrastructure, and other deployment specifics differ.

What demo-first verification catches:

• **Broker incompatibilities**: vendor verifies on IC Markets Raw, buyer deploys on Pepperstone Razor; both are Tier-1 ECN but spread variations and execution differences produce slightly different results. Demo reveals the magnitude of the difference before real capital experiences it • **VPS-specific behaviours**: vendor's verified deployment runs on Beeks Financial Cloud LD4 colocation at 8ms latency; buyer's VPS runs at 30ms latency. Demo exposes how the latency difference affects scalping strategies specifically • **Account-type surprises**: vendor's verification uses a $25,000 account; buyer's first deployment uses $2,000. The smaller account's coarser position sizing introduces friction the verification doesn't show • **Parameter drift between vendor defaults and buyer settings**: vendor verification used specific parameter values; buyer may have inadvertently changed them during installation. Demo catches deviations before they hurt live deployment • **Spread surprises during news events**: vendor's verification doesn't comprehensively expose news-event spread widening on the buyer's broker; demo exposes this in real-time • **Slippage patterns specific to broker**: vendor's broker shows tight, predictable slippage; buyer's broker may have different slippage characteristics that affect specific strategy classes more • **Connection stability issues**: VPS provider reliability, broker connection persistence, EA's ability to recover from disconnections — all exposed during demo run • **Operational workflow development**: buyer develops monitoring routine, alert configuration, journal-keeping practice during demo so the routine is established before real-capital deployment

Demo duration by deployment context:

• **2 weeks minimum**: for any new EA purchase by experienced traders. Catches major incompatibilities and obvious issues • **4 weeks**: for serious capital commitments ($5k+); covers more market regime variation and exposes lower-frequency issues • **8 weeks**: for institutional-tier deployments ($25k+) and for first-ever EA deployments by new traders. Develops the comprehensive operational discipline before real capital is at stake

What demo-first verification doesn't catch:

• **Live execution differences from demo**: many brokers run demo accounts with better conditions than live accounts (tighter spreads, faster fills). Demo provides a baseline but live deployment may exhibit additional friction not visible in demo. After demo, deploy live small first ($500-$2,000) before scaling to full allocation • **Long-term strategy decay**: 2-4 weeks of demo may not expose ML model decay or regime adaptation issues that emerge over months. Demo verification establishes initial fit; ongoing monitoring catches long-term degradation • **Vendor abandonment risk**: demo deployment doesn't verify the vendor will be present 12-18 months later. Vendor longevity vetting is a separate discipline from demo-first verification

Common trader errors that demo-first verification protects against:

• Skipping demo entirely because the vendor's verification looks strong — this is the most common error and produces the most preventable losses • Running demo for too short a period (3-5 days) to catch meaningful issues • Demo-running with different parameter configurations than intended live deployment, so the demo doesn't actually verify the live setup • Demo-running but not reviewing the demo journal critically before live deployment, missing red flags the demo exposed

The editorial position is uniform across all tiers and audiences: demo-first verification is non-negotiable. Even highly-rated EAs from highly-respected vendors require demo verification on the buyer's specific deployment environment. Skipping demo to save 4 weeks is paying that 4 weeks of patience in expected losses from preventable deployment errors. The math is unforgiving.