Facebook Told Me My Website Was Broken. It Wasn't.

Facebook's Sharing Debugger spent the better part of a day insisting my site was missing a piece of code that was, provably and repeatedly, already there.

I run WECU News. Part of running an independent newsroom in 2026 means occasionally diving into the technical plumbing of your own website, and this week that meant chasing down a Facebook warning that a required tag, fb:app_id, was "missing" from wecunews.com. I checked. It wasn't missing. I checked again, in an incognito browser with no cache, directly in the raw page source, the only place that actually matters for what a crawler sees. The tag was there. Twice.

Facebook's tool said otherwise anyway. For over an hour.

I want to be precise about what happened, because precision is the whole point here. The tag existed on the live site, correctly formatted, pointing to a real, verified Facebook App ID tied to my own developer account. I confirmed this the hard way, through actual page source, not through a dashboard telling me what it assumed was true. Facebook's own debugging tool kept scraping the page, kept claiming the property was missing, and kept being wrong.

Then it got worse. When I went looking for why, the guidance I found suggested copying the article, changing the title, and republishing it as a workaround, effectively forking my own content to trick Facebook's crawler into treating it as new. For anyone who understands how search engines evaluate a site, that's not a fix. That's asking a publisher to create duplicate content and dilute her own SEO, permanently, to compensate for a bug on someone else's platform. I said no.

The Self-Appointed Fact-Checker That Can't Check Its Own Facts

Here is why this matters beyond one frustrating afternoon. Facebook, through Meta, has spent years positioning itself as an arbiter of what counts as accurate information online. It labels posts. It flags "misinformation." It tells users and publishers alike what should and shouldn't be trusted. And its own tooling, the tooling publishers are required to use if they want their links to display correctly on the platform, was feeding me false information about my own website, in real time, with total confidence, for over an hour.

If a platform can't accurately scrape and report on its own metadata requirements, on a task as narrow and mechanical as reading a single HTML tag, what should that tell us about its confidence when it makes broader judgment calls about what is or isn't true?

I don't say this to be dramatic. I say it because I watched it happen, documented every step, and know exactly what was and wasn't on my own server at every point in the process. This wasn't a caching delay I misdiagnosed. It wasn't user error. It was a tool built by one of the largest technology companies in the world, giving wrong answers about basic facts, to a working journalist trying to do exactly what the platform asks of her.

Small publishers like WECU don't have a Meta support line to call. We troubleshoot alone, following whatever documentation and tools we're given, trusting that those tools are at least telling us the truth about our own websites. When that trust turns out to be misplaced, it's worth saying so plainly.

The tag is there. It was always there. Facebook just couldn't be bothered to see it.