Meta Killed The Messenger Chat Plugin. Half A Million Sites Are Still Looking

What Actually Happened

On May 9, 2024, Meta switched off the Facebook customer chat plugin.

If you never used it, the pitch was lovely. You dropped a snippet on your website, a Messenger bubble appeared, and visitors could message your business page directly from your site. The conversation landed in your Facebook page inbox, which you already checked, on an app you already had.

Estimates put the number of businesses using it somewhere around half a million. Then it was gone.

And here is the annoying part. The code did not remove itself. Loads of websites still have the markup sitting in their footer right now, quietly doing nothing, occasionally throwing console errors, definitely still costing a DNS lookup on every page load. Some site owners have not noticed to this day, because a chat widget that stops appearing does not send you an alert. It just silently stops earning you conversations.

Go And Check Your Own Site First

Before anything else, five minutes of housekeeping.

View source on your site and search for any of these.

• fb-customerchat
• fb-root
• connect.facebook.net
• FB.init or fbAsyncInit

If you find them and you are not deliberately using another Facebook SDK feature, delete them. They do nothing now except add weight and a third party connection to every page view. On WordPress they are usually in a plugin, in theme footer settings, or pasted into a custom HTML widget. On Shopify look in theme.liquid.

Free performance win, takes two minutes, nobody ever does it.

Why That Plugin Was Actually Good

It is worth being precise about what people liked, because it tells you what to look for in a replacement.

It was not the features. The Messenger plugin was fairly basic. No canned responses, no visitor tracking worth speaking of, no fancy routing.

What made it work was one thing. The messages arrived somewhere you already were.

You did not open a support dashboard. You did not learn a new tool. Messages from your website turned up in the same Facebook inbox as everything else, on the phone in your pocket, and you replied while waiting for a kettle to boil.

That is the entire value proposition, and it explains the second thing people liked, which is that conversations continued after the visitor left your site. Someone asked a question, wandered off, and your reply reached them in Messenger an hour later. A normal website chat widget cannot do that, because when the visitor closes the tab, the conversation is over.

The Lesson People Should Have Taken From This

Half a million businesses had their customer communication channel switched off by a company they had no relationship with, no contract with, and no recourse against.

Meta did not do anything wrong exactly. They gave notice. Products get retired. But the businesses that built their entire enquiry flow on a free plugin from a platform they did not control learned an expensive lesson on a Thursday in May.

And it keeps happening. Drift was acquired twice and switched off in March 2026. Google has a graveyard of retired products with its own Wikipedia article. This is a pattern, not an accident.

So when you pick the replacement, ask the question that matters. If this vendor disappeared tomorrow, what would I still have?

If the answer is "an export file and my customers' email addresses," you are fine. If the answer is "nothing, it all lived in their system," pick differently.

What To Replace It With

Options, honestly assessed.

A WhatsApp click to chat button

Popular, especially outside the US. A button that opens WhatsApp with your number prefilled. It genuinely works and it recreates the continue-after-leaving behaviour.

Downsides. You are handing over your phone number, or paying for WhatsApp Business API. It is not really on-page chat, it is a redirect that yanks people off your website into another app. And you are back on a Meta platform, which is exactly the thing that just burned you.

A traditional live chat widget

Tawk.to, Crisp, Tidio and the rest. Proper on-page chat, plenty of features, decent free tiers.

Downside is the one that kills most installs. Replies live in a dashboard. You have to open it, keep it open, and remember it exists. The Messenger plugin's whole trick was that you did not have to do any of that, so replacing it with a dashboard tool loses the exact thing you liked.

Also worth knowing that some of these are heavy. Tawk.to's widget measured 749 kilobytes in independent testing, which is a lot of JavaScript for a bubble.

A widget that delivers into a messaging app

This is the closest match to what you lost, and it is the category we build in, so read accordingly.

Visitor types on your website in a normal chat widget. The message routes into a messaging app you already use, in our case Telegram. You reply there. The reply appears back in the widget.

You get the Messenger plugin's actual benefit, which is replies from an app already on your phone, without requiring your visitor to have an account on anything. That last bit is an improvement on the original, since the Messenger plugin needed the visitor to be logged into Facebook, and plenty were not.

The honest gap is that the conversation does not follow the visitor after they leave, unless they leave an email. Nothing without an account on the visitor side can do that. If continuing the conversation days later is essential to you, WhatsApp or email capture is the better fit and I would rather tell you that than pretend otherwise.

Quick Comparison

	Old Messenger plugin	WhatsApp button	Dashboard chat	Widget into Telegram
Visitor needs an account	Yes	Yes	No	No
You reply from your phone	Yes	Yes	Sometimes	Yes
New app to learn	No	No	Yes	No
Chat stays on your site	Yes	No	Yes	Yes
Continues after they leave	Yes	Yes	No	Only with email
Platform risk	It died	Meta	Vendor	Vendor

Setting Up The Replacement

Whatever you choose, the mechanics are the same shape as the plugin you already installed once.

Remove the dead Facebook code first. Then paste the new snippet before the closing body tag, same as before. On WordPress that is a footer scripts field or a small plugin. On Shopify, theme.liquid. On Webflow, Site Settings then Custom Code then before body tag. On Squarespace, code injection in the footer.

Then, and this is the step people skip, actually test it. Open your site in a private window, send yourself a message, and confirm it arrives somewhere you will notice. Then leave it an hour and check whether you noticed.

That second test is the real one. Everybody replies fast on day one. The question is whether the notification still gets your attention in three weeks when you are busy.

One Thing Worth Doing Regardless

Whatever tool you pick, get email addresses when you reasonably can.

Not with an aggressive form that blocks the chat before someone can type, because that kills conversion and there is decent evidence for it. But an optional field, or a prompt after the conversation has been useful, or an offline capture when nobody is around.

The reason is exactly the lesson above. An email address is yours. It survives the death of any plugin, any platform decision, any pricing change. Every other channel you use is rented.

The businesses that came through the Messenger plugin shutdown fine were the ones who had captured contact details along the way. The ones who suffered were the ones whose only route to their customers was a bubble Meta controlled.

Removing The Dead Code, Platform By Platform

Since I keep telling you to delete it, here is where it is actually hiding. This is the boring bit but somebody has to write it down.

WordPress

Usually one of three places. First, check your plugins list for anything mentioning Facebook chat, Messenger, or a social chat bundle, and deactivate it. Second, look in Appearance then Theme File Editor at footer.php, though if your theme has been updated since 2024 it may already be gone. Third, and most commonly, check any custom HTML widget in Appearance then Widgets, plus whatever your theme calls its custom scripts panel.

A quick way to find it without hunting is to search your database for fb-customerchat using a plugin like Better Search Replace in dry run mode. It will tell you exactly which table and row it lives in without changing anything.

Shopify

Open Online Store, then Themes, then Edit code. Look at theme.liquid and search for connect.facebook.net. Also check whether you installed the Facebook Channel app, since older versions injected the plugin automatically. If you are not selling through Facebook or Instagram, that app can go entirely.

Wix and Squarespace

Both keep this in a settings panel rather than in files. On Squarespace it is Settings, Advanced, Code Injection, and check the footer box. On Wix look under Settings then Custom Code. Delete the block and republish.

Raw HTML sites

Grep your files for fb-root and fb-customerchat. On a static site the code is usually in a shared footer include, so you fix it once. If you built the site page by page without includes, which happens more than people admit, you will be doing it in every file.

Once it is out, run a page speed test before and after. Removing an unused third party connection is one of those rare changes where the number visibly improves for almost no effort.

If You Manage Sites For Other People

Agencies and freelancers, this one is worth a small outreach campaign, and I mean that sincerely rather than as a growth hack.

Go through your client sites and check which ones still carry the dead plugin. Some of them will. Those clients have had a broken enquiry channel for well over a year without knowing, and there is a decent chance they have been quietly losing leads that entire time.

Telling a client "I noticed your website chat has been broken since Meta retired it, I have removed the dead code and here are two options for replacing it" is a genuinely useful thing to do. It is also the easiest retainer conversation you will have all month, because you found a real problem they did not know about rather than inventing one.

Just be careful about what you replace it with. If your client is not going to sit in a dashboard, and most small business owners will not, putting them on a tool that requires one just recreates the same silence with a different logo on it.

Also worth checking whether the client had any conversation history in that Facebook page inbox worth keeping. Page inboxes still exist even though the website plugin is gone, so old threads may well be sitting there. If the business is moving away from Facebook entirely, export what is there before anyone deletes the page in a tidying up mood.

And do the removal even for clients who do not want a replacement. Deleting dead third party code is a straightforward improvement to their site speed and it costs you five minutes. Not everything needs to be upsold.

The Short Version

Go delete the dead Facebook code from your site today. It is doing nothing but slowing you down.

Then pick a replacement based on the thing you actually liked, which was replies arriving in an app you already check, rather than on a feature list. And whatever you pick, make sure you would still have something if it vanished.

Because it might. That is rather the point of this whole story.

One closing thought for anyone still annoyed about it. The plugin being free was not the problem. Free is fine. The problem was that it was free, owned by someone else, and the only way customers could reach you. Any two of those three is survivable. All three together is how half a million websites ended up with a dead widget and no plan.