Hive Manila

Why our new website scored 77 on launch day

Our new site went live and Google's mobile speed test gave it a 77. The page was not the problem. Two tracking scripts were. Here is what we changed to reach 100.

Developer smiling at a performance chart on his laptop in the evening

We launched the new hivemanila.com just after noon on November 20. A little over an hour later we ran Google’s PageSpeed test on mobile and got a 77.

That stung a little. We had built the site to be fast: no CMS, no database, every page built ahead of time, photos in modern formats. Our own tests before launch scored in the high 90s.

So what changed when the site went live? The tracking went live with it.

The short version, in 30 seconds.

What the test found

The score was dragged down by one number. The main photo on the homepage took 5.2 seconds to appear on a phone. Everything else was fine.

The page itself was light. The weight came from two scripts every business site runs:

  • Google Analytics, about 177 KB, so we can see where visitors come from.
  • The Meta Pixel, about 200 KB, so we can measure Facebook and Instagram ads.

On a fast laptop you would never notice. On a phone with an average connection, the browser was busy downloading and running almost 400 KB of tracking code while it should have been showing the page.

Before launch we test without these scripts, because test visits should not count as real traffic. That is why we missed it.

What we changed

We did not remove the tracking. We made it wait.

The scripts now load on the visitor’s first scroll, tap or key press, or 4 seconds after the page has finished loading, whichever comes first. Anything a visitor does in the meantime is still recorded. The only visits we miss are people who open the page and leave within a few seconds without touching it, and those were not going to become leads.

While we were at it, we fixed three smaller things:

  1. Moving logo strips now pause when they are off screen.
  2. Two small scripts that measured the page while it was still loading now wait a moment.
  3. Photos and fonts are now saved by the browser, so a second visit loads faster.

The result

We ran the same test 14 minutes later.

Mobile Before After
PageSpeed score 77 100
Main photo on screen 5.2 s 1.2 s
Time the page was blocked 160 ms 0 ms

Check yours

Open PageSpeed Insights, paste your homepage and look at the mobile score. If it is below 90, scroll to the list of scripts it flags. On most small business sites in the Philippines, the heaviest ones are not the page at all. They are chat widgets, tracking pixels and plugins added over the years.

You can often fix this yourself by removing what you no longer use and delaying what you still need. If you would rather have it done, our Project Rescue diagnosis finds what is slowing your site and tells you what to fix first.

Common questions

What is a good PageSpeed score?

90 or above on mobile is good. Between 50 and 89 needs work, and below 50 is poor. Test the mobile score, since that is how most of your visitors arrive.

Does Google Analytics slow down a website?

It can, a little. The tracking script and others like the Meta Pixel compete with the page on slow phones. Loading them after the first scroll or tap keeps the data and removes most of the delay.

Does website speed affect Google rankings?

Yes, as one of many signals. More importantly, slow pages lose visitors before they see anything, which costs you enquiries whatever your ranking.

Get the Sunday read

One short email a week with our newest insights and case studies. Sunday, 8 AM Manila time.

Is your site slow on a phone?

A short brief, then a written proposal within 1 working day. We talk once the work needs it.