Why Visible-Text Keyword Detection Misses Hidden Data and How to Fix It

You set up keyword monitoring, you can literally see the word sitting in the page's HTML when you open DevTools — and yet the alert never fires. Nothing is broken. Standard keyword detection reads the visible text of a page: the words a human can actually read on screen. If the value you are watching lives in a hidden element, a data- attribute, or an inline script, it is real data but it is not visible text, so ordinary detection walks right past it.

The fix is one checkbox. On the Detect Keyword tab of the Auto Refresh extension by Extfy, enable Find monitor text in source and the keyword search switches from the rendered text to the complete HTML source — hidden elements, attributes, and scripts included.

Auto Refresh

English 

Running 

Time Interval

Refresh List

Detect Keyword

Enter your tags

in_stock ✕

Notification & Highlight Settings for Keyword

 Notify the user 

When keyword found or selector clicked

When keyword not found or no selector matched

 Play sound on keyword found/not found 

 Find monitor text in source 

Save

Where the hidden data actually lives

Modern pages carry far more information in their markup than they ever show on screen. The keyword you care about is often sitting in one of these places:

• Hidden elements — a status message wrapped in display:none or visibility:hidden, waiting for a script to reveal it.
• data- attributes — values like data-availability="in_stock" or data-status="approved" that drive the UI but never render as text.
• Inline scripts and JSON — state objects the page ships in a <script> block and only renders later with JavaScript.
• Attributes like title, alt, and aria-label — readable by machines and screen readers, invisible to text matching.
• HTML comments and template blocks — markup the developers left behind that still reflects the real state of the page.

Why normal keyword detection cannot see any of it

Visible-text detection is deliberately built to match your intuition: it reads the page the way a person does, after the browser has rendered everything, and it only considers words that actually appear on screen. That keeps it fast and keeps the results clean — a match means a human could see the word too. The cost of that design is a blind spot. Anything that exists only in the markup is, by definition, not visible text, so the detector never even looks at it. The page can be full of the exact string you typed as your keyword and detection will honestly report “not found,” because in the rendered view it genuinely is not there.

The fix: turn on Find monitor text in source

1. Open the Auto Refresh extension on the page you are monitoring and switch to the Detect Keyword tab.
2. Enter the keyword exactly as it appears in the source — for example in_stock, not the “In stock” label the page might display later.
3. Tick Find monitor text in source.
4. Click Save. From the next reload onward, every check scans the complete HTML source — hidden elements and scripts included — instead of only the visible text.

Visible-text vs. source detection at a glance

• What it reads: visible-text mode reads the rendered page a person sees; source mode reads the raw HTML the server sent — every tag, attribute, comment, and inline script.
• Speed: visible-text mode is lighter; scanning the entire source does more work on each reload, so expect checks to be a little heavier.
• False positives: visible-text mode rarely misfires; source mode can match a keyword in background markup that has nothing to do with what is on screen.
• Best for: visible-text mode for everyday monitoring; source mode for developers and power users watching values the page stores but does not show.

A real example: catching the change before the page shows it

Picture a product page that displays “Out of stock” to shoppers while the markup already carries data-availability="in_stock" — the data has updated, the UI just has not caught up. Visible-text detection keeps matching the stale “Out of stock” label. With Find monitor text in source enabled and the keyword set to in_stock, Auto Refresh spots the change the moment it lands in the HTML, often a beat before the visible page updates. The same trick works for hidden error flags, approval states in admin tools, and any value a page renders late with JavaScript.

Keep source matching accurate

• Be specific. Short, common words appear all over a page's markup by accident; a distinctive string like in_stock or status-approved is far safer than stock.
• Watch for phantom matches. If alerts fire when nothing visibly changed, your keyword is matching background markup — tighten the keyword or switch that rule back to visible-text mode.
• Enable it per rule, not everywhere. If standard detection already catches what you need on a page, leave this option off there for faster, cleaner checks.

Visible-text detection is the right default — until the thing you are watching never actually renders. When the keyword lives in a hidden element, an attribute, or a script, flip on Find monitor text in source on the Detect Keyword tab and Auto Refresh reads the page the way DevTools does, catching the data that normal detection quietly misses.