“Self-hosted platforms are slower than the big SaaS ones.” On the operations that actually hurt, a new set of load tests says it’s backwards. On the long, repetitive jobs that quietly punish a classic e-commerce stack — full catalogue sync, bulk re-indexing, feed reconciliation — a checkout-first platform reportedly ran about ten times faster than a conventional baseline.

What was measured

The benchmark targets the jobs nobody demos but everybody runs in production. As the analysis on VBWD describes it:

In VBWD’s own load tests, the long, repetitive jobs that quietly punish a classic e-commerce stack — full catalogue synchronisation, bulk re-indexing, feed reconciliation — ran roughly 10 times faster than a conventional catalogue-first baseline.

The reasons are architectural, not a tuned hot path: the catalogue isn’t the spine (the point of truth is the invoice line item, so a sync touches far less per record), and an event-driven core lets a change propagate as a targeted event instead of a wholesale re-computation.

Why 10x is money, not seconds

The write-up is blunt about where the saving lands:

Long jobs are where cloud spend concentrates

— CPU-minutes, database load, the oversized instance kept “just for the nightly sync.” It’s also honest about the limits:

This is a benchmark of specific long-running operations, not a blanket claim that every request is 10x faster.

Results vary by workload and hardware, and a single page load won’t feel different. But on the heavy jobs that drive an infrastructure bill, designing around the invoice and the event bus pays off by roughly an order of magnitude.

The full methodology is on VBWD, and the platform is source-available on GitHub if you want to run the numbers yourself.

Written for Red Robot with AI assistance and human editing. Adapted from an original analysis published on VBWD.