Performance comparison results

We closely monitor the performance of once.to to provide our users with the best possible experience.

What we are particularly proud of is the unparallelled performance of our redirection engine, which is so thoroughly optimized that no other service can match our throughput figures.

Here are our latest measurements, including detailed reports1:

Service/Metriconce.to💪bit.lyt.lypolr.meshorturl.atbl.ink
Throughput, requests/s14261051354320788
Average single request duration, ms67.9191.1528102320473.91080
Average request duration (in burst), ms0.7010.95128.57123.3284.82811.302
Overall test duration, s28.038210210207452
ReportReportReportReportReportReport

Graphs

Below you can see the two most important performance metrics plotted on a chart:

Throughput chart (more is better).
Throughput chart (more is better).
Average burst request duration chart (less is better).
Average burst request duration chart (less is better).

Why is that important?

As you might have heard, search engines tend to punish sluggish websites by lowering them in their search results. And rightly so: according to stats by Google, about a half (!) of visits get abandoned if a website takes longer than three seconds to open:

The percentage of abandoned visits for sites taking longer than 3 seconds to load. Source: Google.
The percentage of abandoned visits for sites taking longer than 3 seconds to load. Source: Google.

So, are you still considering using t.ly, which takes 2.81 s on average only to redirect to your website?


  1. Sampled on 40K HTTPS requests to two different URLs, emulating up to 100 simultaneous users. t.ly and polr.me could not complete all the 40K requests in the test within the time limit of 210 seconds. ↩︎


Tags: , , , , ,