Digital Experience: Latency Budget Allocator

Pick a digital experience — checkout, AI chat, live video — and allocate its latency budget across origin, regional, edge, and client tiers. The cost index shows what you pay for each millisecond.

— —

Origin compute —

Cheapest per ms, slowest from London-egress. Where heavy logic and writes still live.

Regional edge —

Stateful logic close to the user. Personalisation, auth, A/B, rate-limit, model inference.

Last-mile edge —

POPs and CDN compute. Cache, transforms, light edge functions, rule-based routing.

Client compute —

Free in unit cost, but you pay in trust, battery, and device variance.

Total latency —

Budget headroom —

Cost index — 100 = origin-only baseline

Content-era baseline — 90% origin, 10% client

—

Read the full From Content Wars to Experience Wars insight →