🔎 Focus: Rendering
🔴 Impact: High
🟡 Difficulty: Medium-low

1. Your website was “Google enough“

For a decade, your website was "Google enough." Users and Google got quite sophisticated and tolerate "slow" sites or heavy JavaScript.

That era is over.

With "AI search" we might need to pay attention to Technical SEO deserves.

With new AI crawlers in place like ChatGPT-User, bad performing websites won't be as visible on LLMs as they were on Google.

2. Faster website = More AI visibility

Let’s not even get into the fact that faster websites convert better. Let’s just focus on client acquisition for now.

Faster websites are in around 1.5x more cited by LLMs than their “slower“ counterparts. Faster website are more visible. More impressions > traffic > conversions.

In reality, is not site speed what matters but how “light” each page is. Think about just a pure text page, no CSS, no Javascript. Just plain text. That would be the easiest for AI Crawlers, like OAI-SearchBot (Open AI’s), to extract content from.

Javascript Is The Enemy: It Hides Valuable Content

For example, websites that rely on React, Angular or Vue (Javascript frameworks) are vulnerable to being left out AI responses: the content is invisible to AI crawlers.

🧑‍💻 Technical Note: Even thought Google has capabilities to execute Javascript, is best practice to use as much Server-Rendered content as possible to avoid any delays in indexing.

The Gemini Stress Test - Try this at home

We are testing the website halfords.com, which depends a lot on Javascript rendered HTML. [You can do the same with you own site].

We will test a product page and see if we can get the reviews of the product using the real time browser of Gemini. Prompt this:

browse and extract the reviews of this product URL:
https://www.halfords.com/motoring/batteries/car-batteries/yuasa-ybx7335-12v-80ah-780a-efb-start%2Fstop-battery-485454.html

This will push the Gemini Live Browser tool to read your page: no JS, only plain HTML text.

Usually the reviews are loaded with Javascript therefore not available on the initial rendered HTML. We are trying to retrieve the content from one specific block (an actual user review):

🧑‍💻 Note: reviews are a fantastic method of unique content and trust signals. Having that content available will be essential for AI search visibility as it will touch extremely specific user pain points.

The response of Gemini, based on the Browser tool, is:

The Gemini Browser is acting like a simple scraper.

🧑‍💻 Technical Note: Gemini is able to render Javascript as it relies on Google’s infrastructure. But even then, with this example we see that more complex Javascript tasks (nested) might fall away their read capabilities.

Rendering Gap = Invisible Content

If Gemini returns "No reviews found," but you see them on your screen, you have a Rendering Gap. This rendering content gap is currently costing you citations in AI-driven answers.

LLMs won’t be able to retrieve that piece neither offer it to users via citations.
Google crawlers will have to evaluate your URLs multiple times to make sure it is crawler all the way = more expensive for Google to crawl you

3. Fix your AI visibility with Technical SEO

Aim to have a light website. It should render fast and provide all content straight from the server. Use Javascript mostly for enhanced user experience but do not use it to render content, of any kind.

Technical Tip	Action
1. Server-Side Rendering (SSR) / Prerendering	All content should be present in the initial HTML (even internal links, recommended pages)
2. Use Clean HTML	Employ semantic HTML tags (<article>, <nav>, <h1>, <p>) for clear content hierarchy.
3. Improve Core Web Vitals	Faster load times improve the site's crawl budget, allowing bots to process more pages efficiently.