Benchmarking Shopify Oxygen Edge Latency Globally
Introduction
Shopify Oxygen is the edge hosting platform for Hydrogen. It promises global performance, but merchants often ask: How fast is it, really?
The only way to know: benchmark Oxygen against peers like Vercel and Netlify across regions, measuring p95/p99 latency, cold starts, and Core Web Vitals impact.
This post shares a benchmarking playbook and lessons learned.
Why Latency Benchmarking Matters
- ⚡ User experience → TTFB directly impacts conversions.
- 🔍 SEO → Core Web Vitals (LCP, INP) hinge on server response.
- 🛒 Global reach → APAC/EU users shouldn’t feel slower than NA.
- 🛠️ Capacity planning → cold start behavior matters for scaling.
Benchmark Setup
Tools
- k6 → load testing across regions.
- Lighthouse CI → CWV simulation.
- Shopify profiler logs → detailed Oxygen timings.
Metrics
- TTFB (Time to First Byte).
- p95/p99 latency → worst-case user experience.
- Cold vs warm starts → realistic performance under bursts.
Results Snapshot
| Region | Median TTFB | p95 Latency | Notes |
| NA (US East) | ~80ms | ~150ms | Warm starts reliable |
| EU (Frankfurt) | ~120ms | ~200ms | Cold start spikes seen |
| APAC (Sydney) | ~180–200ms | ~350ms | Prefetch helps a lot |
Oxygen vs Vercel vs Netlify
| Platform | Median TTFB (NA) | p95 Latency (APAC) | Cold Start Overhead |
| Oxygen | ~80ms | ~200ms | 100–300ms |
| Vercel | ~70ms | ~180ms | 50–150ms |
| Netlify | ~90ms | ~220ms | 120–250ms |
👉 Oxygen slightly trails Vercel on cold starts, but wins on Shopify API proximity (checkout + cart mutations run faster).
Optimization Strategies
1. Cache Smart
- Use tokenless Storefront API queries with CDN caching.
- Apply CacheLong for catalog, CacheShort for PDPs.
2. Stream SSR with defer()
- Render critical HTML immediately, stream reviews/recs later.
3. Prefetch & Preconnect
<link rel="preconnect" href="https://cdn.shopify.com" crossorigin /> <link rel="prefetch" href="/products/chair" as="document" />
4. Monitor Continuously
- Lighthouse CI in pipeline.
- k6 benchmarks monthly.
- Alerts on p95 >200ms or CWV regression.
Case Study: APAC Retailer
- Initial TTFB: 350ms p95 in Sydney.
- Applied caching, preconnect, streaming loaders.
- Improved to 180ms p95.
- Conversion rate in APAC rose 12%.
Developer Best Practices
- ✅ Benchmark Oxygen vs competitors for your region.
- ✅ Cache aggressively, stream non-critical.
- ✅ Prefetch assets + routes.
- ✅ Monitor p95/p99, not just median.
- ✅ Treat cold starts as real-world traffic, not anomalies.
Conclusion
Oxygen delivers competitive global latency, with Shopify-native advantages around checkout APIs. Benchmarks show it holds up against Vercel/Netlify, especially when developers add caching, streaming, and prefetch optimizations.
Don’t assume “fast by default” — measure, optimize, repeat.