Why Your Link Preview Shows the Wrong Image (and How to Fix It)

The tags that build the preview

<meta property="og:title" content="How to Center a Div">
<meta property="og:description" content="Five ways, and when to use each.">
<meta property="og:image" content="https://example.com/og/center-a-div.png">
<meta property="og:url" content="https://example.com/guides/center-a-div">
<meta property="og:type" content="article">
<meta name="twitter:card" content="summary_large_image">

These are Open Graph tags, created by Facebook and now read by almost every platform: LinkedIn, WhatsApp, Slack, Discord, iMessage. X reads its own twitter: tags first and falls back to Open Graph for anything missing, so the one tag it really needs is twitter:card set to summary_large_image for a big image.

Why the image is wrong or missing

• The image URL is relative. og:image must be a full URL including https and the domain. A path like /og.png often shows nothing.
• There is no og:image, so the platform guesses, often picking a logo or the first image it finds.
• The image is the wrong shape and gets cropped. Wide previews use roughly 1.91 to 1, which is why 1200 by 630 pixels is the standard size.
• The image is too small. Below about 600 pixels wide, many platforms fall back to a small square thumbnail.
• The image is blocked: behind a login, refused by robots.txt, or too large a file for the platform to fetch quickly.
• Every page shares one site-wide image, so every link looks the same. Give important pages their own.

The fix that seems not to work: caching

Platforms read your tags the first time a link is shared and then reuse that preview, sometimes for weeks. So you fix the tags, share again, and see the old preview. The tags are fine; the platform simply has not looked again.

• Facebook and WhatsApp: use Meta's Sharing Debugger and choose to scrape the URL again.
• LinkedIn: use its Post Inspector, which fetches the page fresh.
• Elsewhere: share the URL with a harmless query string added, such as ?v=2. To the platform it is a new URL, so it reads the tags again.

Check what the crawler actually sees

If your site builds its page with JavaScript in the browser, the tags may not exist in the HTML the platform downloads, because preview crawlers generally do not run JavaScript. View the page source, not the inspector, and confirm the tags are there in the raw HTML.