Home > Programming, Servers & Scripts >

Intial Load or Subsequent Load Test? (5)


03-17-2016 09:59 PM #1 glennb (Member)
Intial Load or Subsequent Load Test?

Hey,

I'm currently running pop traffic.

My question is, when you test your landers (pingdom or gtmetrix) do you test for the initial load time or the subsequent load time?

When I first test my optimized landers, my initial load time is around 2sec - but I rerun the test and it's down to 300-500ms.

For pop traffic, should I keep optimizing to reduce the initial load time to under 500ms?


03-17-2016 10:44 PM #2 ysekse (Member)

Also wondering about this.


03-18-2016 12:19 PM #3 caurmen (Administrator)

Subsequent load.

You're running with a CDN, I assume? If so then the first load is not at all representative, because the CDN has to fetch the page from the origin server. Subsequent load times are much more representative of what it'll look like.

This does to a certain extent depend on how often the CDN refreshes its copies of pages, and what volume you're running. But on pops unless you're running tiny volume that shouldn't be an issue.

Under 500ms is probably at the point where there are other things you can optimise for higher gain, at least until the campaign's solidly profitable.

Hope that helps!


03-18-2016 01:43 PM #4 glennb (Member)

That helps! Thanks caurmen


11-15-2017 11:52 AM #5 vortex (Senior Moderator)

A follow-up discussion on the same topic:

https://stmforum.com/forum/showthrea...l=1#post330593



Amy


Home > Programming, Servers & Scripts >