INP + other speed measures & DOM Size & Content-visibility

**chaumi** · 1 February 2024, 7:07 am

Google (reportedly) kicks off INP as a CWV measure on 12th March.

https://web.dev/blog/inp-cwv-march-12?

So, 1. not specifically related to INP given we don't know how G will act on it yet and 2. given INP is just one measure and we know that many measures interact/have impacts on each other and 3. there are multiple potential root causes for all loading/interaction speed-related measures...

One of the often cited (possible) reasons for dropping a few points in overall page speed scores is an excessive DOM size (and I expect this will also be the case with direct reference to improving INP).

'Easy' to get round, right? Just cut a bunch of content/code/elements from the page. But that's not always practical and may stroll into 'shoot yourself in the foot' territory.

Hence, some questions:

1. Anyone have any experience of deliberately pruning a well-ranked page with an excessive DOM size, to reduce the DOM significantly enough without negatively impacting rankings/impressions (and maybe even seeing a ranking/impressions improvement)?

2. Anyone know any (straightforward to implement) ways of reducing DOM without actual pruning of content?

3. Anyone looked at this new CSS solution yet (content-visibility), that appears to lazy load the entire (written) content of a page, hence effectively eliminating the negative effects of a large DOM?

4. Does anyone think that it can be implemented easily (by a lesser-skilled 'coder' who can just about battle through understanding it, but still with heavy limitations in the understanding of how to make it work)?

https://web.dev/articles/content-visibility

**bpmee** · 1 February 2024, 11:02 am

Originally Posted by chaumi

'Easy' to get round, right? Just cut a bunch of content/code/elements from the page. But that's not always practical and may stroll into 'shoot yourself in the foot' territory.

I would be very careful about removing anything that potentially benefits users more than a PS score. Of course, if the INP is bothering users, then absolutely remove it.

On the other hand, it would be better to wait if a clunky but useful table has earned a link or two, and the average Joe doesn't mind a 250ms wait toggling through data.

As you noted, I agree that DOM size and other factors could be a way to guess INP even if INP didn't exist.

FTR, some of the biggest sites in the industry have below-average PS performance scores and rarely get dinged for it. Authority status confers the right to add a lot of bloat to pages that may or may not help the user, in favor of looking good.

Originally Posted by chaumi

1. Anyone have any experience of deliberately pruning a well-ranked page with an excessive DOM size, to reduce the DOM significantly enough without negatively impacting rankings/impressions (and maybe even seeing a ranking/impressions improvement)?

So, see above rationale. This will be interesting to watch in the coming months.

Originally Posted by chaumi

2. Anyone know any (straightforward to implement) ways of reducing DOM without actual pruning of content?

1. Remove any redundant <div> or <span> elements, particularly if their sole purpose is visual formatting. Try retooling your CSS instead. I need to do this myself at some point.

2. Reduce page content to the bare minimum. For example, WP category pages show 5 articles instead of 15 or 20.

A HTML table that displays a year's worth of data (default) could instead show 3 months. Add a button that allows the user to select more on their own.

You could also setup a JS widget to display heavy HTML table data, like an odds feed or box scores. However, Google can't index JS tables for featured snippets. I see some sites use basic HTML tables just to get the occasional featured result. Be cautious here.

Originally Posted by chaumi

3. Anyone looked at this new CSS solution yet (content-visibility), that appears to lazy load the entire (written) content of a page, hence effectively eliminating the negative effects of a large DOM?

Thanks for sharing this, I haven't tried it yet. CSS is my weak spot. I'd be careful to test thoroughly test this, just in case crawlers mistake it for content that should be treated as "invisible" and thus not important to users.

Originally Posted by chaumi

4. Does anyone think that it can be implemented easily (by a lesser-skilled 'coder' who can just about battle through understanding it, but still with heavy limitations in the understanding of how to make it work)?

Possibly if you're just reworking and minimizing HTML without the need to invent anything new.

Personally, I'm going to wait a couple weeks after the March rollout to see what, if anything, happens.

Parting note: sometimes a slow INP can be attributed to someone with a slow connection. For example, someone in Australia hitting a Canadian server with an AJAX request to see a list of 2008's top 100 Champions League goal scorers presented in a multi-column HTML table. While it would be quick in Toronto, such a page would likely lag in Sydney.

Prioritize the average, "target" customer over the aberration that might be affecting your result.

**bpmee** · 1 February 2024, 11:27 am

And another: Check if any WP plugins are affecting your INP scores. If so, edit or replace them with something faster. For example, social media plugins that display data "on click" might be dulling your INP results.

Worth a mention before people start tearing their site to pieces only to realize someone else's code is the problem.

**chaumi** · 1 February 2024, 12:31 pm

Thx bpmee. Some great suggestions.

I think the big(ger) point I was making with all this is around that new content-visibility CSS solution....because if that works, and it's doable, and it doesn't cause other issues, it (theoretically) 'solves' a whole bunch of page speed related issues without actually going through the grief of having to truly solve them.

And you're right, of course, whether we (or some of us) do actually have to solve them anyway depends on what G really does regarding INP. It might be just another scare tactic to 'encourage' more sites to speed up, but with no or little background intent to 'punish' anyone that doesn't.

Personally I believe they will step up this time with INP and we'll find that sites not meeting the requirements will see some negative consequences. But that's guesswork, and to be confirmed or ruled out over the next 6-12 months.

**bpmee** · 1 February 2024, 1:05 pm

Originally Posted by chaumi

Thx bpmee. Some great suggestions.

I think the big(ger) point I was making with all this is around that new content-visibility CSS solution....because if that works, and it's doable, and it doesn't cause other issues, it (theoretically) 'solves' a whole bunch of page speed related issues without actually going through the grief of having to truly solve them.

Absolutely on the content-visibility CSS fix!

**allfreechips** · 1 February 2024, 3:36 pm

also look at structure see if you can eliminate nested elements like <div><div><div><span> ext...

Thread: INP + other speed measures & DOM Size & Content-visibility

LinkBack

Thread Tools

Search Thread

Display

INP + other speed measures & DOM Size & Content-visibility

The Following User Says Thank You to bpmee For This Useful Post:

The Following User Says Thank You to chaumi For This Useful Post:

Posting Permissions