A communal belief among galore SEOs is that Google’s Web Rendering Service has a five-second rendering model and that thing that happens aft that five-second model won’t beryllium included successful the type of the page that Google indexes.
The thought started pinch thing Google’s Martin Splitt said connected astatine slightest 2 occasions and yet became wide accepted arsenic true. Dave Smart of Tame the Bots tested the mentation and discovered not only that the five-second model isn’t true, but besides that Google’s Web Rendering Service tin region and restart the rendering clock.
So who’s right, the SEOs who guidelines their opinions connected what Martin Splitt said, aliases Dave Smart, an SEO successful the UK? My money is connected Dave Smart, and you’ll soon understand why.
What SEOs Say
The thought promoted successful galore SEO blogs is that there’s a five-second model that is the magnitude of clip that Google’s Web Rendering Service (WRS) waits earlier capturing the rendered Document Object Model (DOM) of a web page.
Google’s Web Rendering Service is simply a postulation of services that do six things:
- It receives the HTML that Googlebot crawled.
- Fetches the web page’s resources, specified arsenic CSS, JavaScript, and images.
- Loads the page successful a headless Chromium browser.
- It adjacent executes the page’s JavaScript.
- And past produces a snapshot of the page’s DOM.
- At this constituent it passes that DOM snapshot into Google’s indexing system.
SEOs Have Two Variations Of Five-Second Window Theory
1. Some SEOs opportunity that the rendering process wrong Google’s Web Rendering Service has astir 5 seconds to nutrient the DOM that Google snapshots. They opportunity location is simply a five-second play during which the page is loaded, JavaScript is executed, and Google captures a snapshot of the rendered DOM for indexing.
2. The 2nd variety is that important contented has to look successful the DOM wrong 5 seconds because that’s erstwhile Google is expected to seizure the DOM snapshot.
Some SEOs moreover declare to person tally a trial and opportunity they person confirmed that location is so a five-second window.
What Did Martin Splitt Say?
Martin Splitt did astatine slightest 2 video presentations wrong the past 7 years successful which he introduced the SEO world to the conception of a five-second rendering window.
Here’s what Splitt really said astir that five-second model (at astir the 18 infinitesimal people successful the video):
“So what do you request to know, what do you request to return distant from this?
All websites get rendered nary matter if they incorporate JavaScript aliases not.
What we spot aft the rendering, that’s the DOM snapshot that will beryllium taken into indexing. That’s what you attraction for.
Not the screenshot, not immoderate weird cache connected the hunt results. Don’t usage that.
Don’t usage for illustration position source. View root doesn’t springiness you DOM information. Use the testing devices that we provide. They’re really giving you what came retired of rendering.
In median, a page’s rendering queued for 5 seconds.
So that intends earlier 1 of these beautiful ones picks them up, they’re connected median 5 seconds successful there. 90th percentile is simply a fewer minutes. So we’re talking minutes, not weeks aliases months.
Rendering aggressively caches resources, truthful you don’t request to interest excessively overmuch astir your crawl fund regarding rendering. It doesn’t make that overmuch of a difference.
And if you want to trial your stuff, do not do weird stuff. Just usage Google Search Console’s Inspect URL. It shows you what happens successful rendering, OK?
Don’t beryllium acrophobic of rendering aliases JavaScript.”
It’s clear that Splitt is not describing a five-second limit for the Web Rendering Service. He’s describing really agelong pages wait, connected average, successful the rendering queue earlier rendering begins. Once a renderer picks up the page, location is thing said astir a clip limit for rendering.
It’s clear that there’s a disconnect betwixt what SEOs deliberation that Splitt said and what he really said.
Dave Smart’s Web Rendering Services Test
Dave Smart (LinkedIn profile) ran an research to trial the five-second model theory. He created a trial page that intentionally delayed contented while signaling aggregate timing measurements to find whether Google’s Web Rendering Service (WRS) really did extremity rendering aft 5 seconds.
The measurement the trial worked was he created a trial page that made POST requests to a PHP script. That PHP book delayed its consequence by a random 3 to six seconds earlier serving content. Smart utilized 2 of these API calls successful the test, truthful together they could return betwixt six and 12 seconds to complete.
What The WRS Test Discovered
Smart’s trial initially was astir testing the five-second model theory, but he ended up pinch a 2nd uncovering that was wholly unexpected.
At first, the research appeared to corroborate the five-second model mentation because the JavaScript timer ran for 5 seconds, giving the belief that Google’s Web Rendering Service had a five-second rendering limit.
He wrote:
“Looking astatine the screenshots, you tin spot the setInterval() loop ran for 5 seconds, and updated the title element. …So, a logical conclusion would beryllium that the WRS has a 5 2nd render limit, right?”
But different measurement contradicted that conclusion. The page besides made server-side API calls that took betwixt six and 12 seconds to complete, yet Google’s Web Rendering Service still waited for the delayed contented and included it successful the rendered DOM.
Smart recovered what seemed to beryllium a paradox, which begged the question: How tin Google’s Web Rendering Service hold longer than 5 seconds if the JavaScript timer only precocious 5 seconds?
The reply is that Google’s Web Rendering Service uses a virtual timepiece alternatively of relying connected existent elapsed time. While waiting for web requests (such arsenic server-side API calls), it tin region that virtual clock, allowing much real-world clip to walk than the JavaScript timer reports.
Smart explained:
“The WRS does clip successful its ain way. Rather than the hardware clock, for illustration you person connected your machine aliases server, the WRS uses a virtual timepiece that they tin control. This intends that they tin velocity up, slow down, aliases moreover region clip for the headless Chrome lawsuit if they choose.
How do we cognize that 5 seconds is not a limit? Remember those api calls that were made? These are delayed server side, for betwixt 3 and 6 seconds. So the full clip for some api calls could beryllium anyplace betwixt 6 and 12 seconds, longer than the celebrated 5 2nd limit.”
We Have A Better Understanding Of Web Rendering
The SEO organization owes Dave Smart a connection of acknowledgment for running this clever experiment and confirming that location is nary five-second clip limit successful narration to Google’s Web Rendering Service. Given that Martin Splitt’s comments besides align pinch Smart’s trial results, I deliberation we tin confidently adjacent the book connected the story of the five-second render limit.
Watch Martin Splitt At About The 18 Minute Mark
Featured Image by Shutterstock/Cristina Conti
English (US) · 
Indonesian (ID) · 
