How OzSnow built infrastructure that holds 100% uptime through its busiest season.
Industry
Entertainment & Leisure
Company size
10 - 50 Employees
About
OzSnow is an Australian-owned and operated tour operator and resort owner specialising in affordable, all-inclusive snow holiday packages for skiing and snowboarding, catering heavily to beginners, students, groups, and families by bundling transportation, accommodation, lift passes, and gear hire in one place.
"Halcrow took over an environment that was fragile and opaque, and made it resilient fast. We went from peak-season outages and failed checkouts to a full season of 100% uptime, with clear monitoring and alerts so we were never surprised."
100% uptime
Through the full ski season, up from frequent crashes during traffic surges
~40% lower
Monthly AWS spend, achieved by replacing a single over-provisioned server with infrastructure that scales automatically with demand
The Situation
OzSnow is a one-stop-shop travel operator for snow trips, sending customers to resorts across Australia and Japan, with a business model built around a short, intense winter peak season. As OzSnow continued to grow, the team saw an opportunity to modernise the infrastructure behind their 18 global websites — building a platform genuinely ready to perform at its best during the moments that matter most.
The Key Result Metric: 100% uptime during peak ski season, ensuring every customer could book with confidence during OzSnow’s most important sales window.
Why they called us
OzSnow’s leadership wanted a partner who could bring genuine infrastructure ownership and accountability to the platform — someone who would take real responsibility for how the system performed under pressure, not just provide advice from a distance.
Halcrow was brought in to take complete ownership of the infrastructure environment, with a clear mandate: build a platform that performs beautifully exactly when OzSnow’s biggest campaigns and busiest season hit hardest.
How we worked
Infrastructure ownership from day one
We took complete ownership of the infrastructure environment, with direct access to the AWS account, the server instances, and the monitoring layer underneath everything.
We began with visibility, introducing real-time performance monitoring across all 18 sites — giving the team a clear, detailed picture of exactly how the platform behaved under real load.
Visibility Has to Come Before Action
Putting strong monitoring in place first meant every decision that followed was grounded in real data, building real confidence in the platform’s direction.
Recovery, built in four phases
Phase 1: Architecture Redesign (Weeks 1–3)
We moved from a single shared server to a fully isolated environment for each individual application — a standard, proven infrastructure practice that gave every site its own resilient foundation.
Phase 2: Auto-Scaling Implementation (Weeks 3–5)
We configured the platform to expand and contract automatically with real demand, ensuring OzSnow’s biggest marketing campaigns would always be met with the capacity to match. This also delivered a strong cost benefit: monthly AWS spend dropped by roughly 40%, since resources now scale precisely with real demand rather than running at peak capacity year-round.
Phase 3: Traffic Routing Simplification (Weeks 5–7)
We simplified the entire system down to a single, unified traffic gateway — easier to manage, with a much faster path to diagnosing anything that needs attention.
Phase 4: Monitoring and Alerting (Ongoing)
With alert thresholds configured for every site, OzSnow is now notified within 15 minutes of anything that could affect a customer’s experience — giving the team genuine peace of mind, season after season.
WHAT CHANGED
OzSnow now runs on isolated, auto-scaling environments for every site, with real-time monitoring across the entire platform. Peak season uptime reached a full 100% through the entire ski season, with every site properly isolated from the others. Monthly AWS cost dropped roughly 40%, and incident response moved from reactive to fully proactive — catching issues at a defined threshold well before they could ever affect a customer.
The result that mattered most: OzSnow went through an entire peak ski season (their whole annual revenue window) with zero infrastructure outages during any major traffic period.
WHY THIS WORKED
The single point of failure pattern
Seasonal businesses like OzSnow face a genuine infrastructure stress test every year. Designing specifically for that peak (rather than for an average day) is what gives a platform like this real, dependable resilience.
Visibility First, Then Action
Starting with strong monitoring meant every architectural decision that followed was grounded in real data about how the platform actually behaves under load.
A pattern worth repeating
Auto-scaling exists precisely for businesses like OzSnow, and it’s neither expensive nor technically complex to implement well — it just takes genuine ownership and a clear focus on the moments that matter most.
what you're buying
If you want your platform to perform beautifully exactly when you need it most, this is what genuine infrastructure ownership looks like — isolation, auto-scaling, and simplified traffic routing, built around your business’s real seasonal rhythm.
Ready to make your most important sales periods your most reliable ones? Book a 20-minute call with Sam Halcrow on 0431197004 or sam@halcrow.com.au.