Why your updates feed went quiet — and what it actually costs

Before you read the rest of this, go and open your own updates feed — not the dashboard, the page a customer sees — and find the date on the newest post.

On one panel feed we have been watching, the most recent entry is from April 2025. Nobody decided to stop. There was no meeting about winding down announcements. It went quiet the way every one of these goes quiet.

Feeds die because publishing is a second place to type

Here is the shape of it. A service’s price goes up. You change it in the panel, which takes four seconds, because that is where a price lives. Then, to tell anybody, you open a different tool, log in, choose a category, write a title, write a body, and publish. The change went live for customers before you started typing. The post is now a chore with no consequence attached to skipping it — and a chore with no consequence attached gets skipped on the first busy day and is never resumed.

Look at the categories on almost any panel’s feed and the real problem is right there in the labels: new service, price increased, price decreased, service disabled. Every one of those is a row in a database you already own. The announcements tool does not know what a service is, so it asks a human to retype a fact the panel is already holding. That is why feeds go stale. It is not discipline.

What a quiet feed costs

Not brand damage, which is the way this is usually sold. Tickets. A price a customer did not expect, a service that disappeared from the list, an order attempted against something that was disabled that morning — each of those arrives in the inbox as a question, and each one costs a reply. On a panel, an updates feed is not a marketing surface. It is deflection, and the way to judge it is by what stops arriving.

The prices, as they actually read

All of these came off the vendors’ own pricing pages on 5 September 2026. Comparison blogs are stale on at least one of them — AnnounceKit is widely listed at a price it does not charge — so check the vendor page yourself before you sign anything.

The unit you are billed in is the actual problem

Beamer’s free tier is 1,000 MAU, and a MAU is defined as a unique user who loads any page the script sits on. Not a customer — a visitor. A panel is an SEO business: most of the traffic on your public pages is logged out and will never place an order. So the bill tracks the number with the least connection to your revenue, and past the allowance it is another $50 a month for each additional 5,000 of them.

Set that against what the platform underneath costs. Perfect Panel’s entry plan is $50 a month, and that $50 is the panel, the order automation, the ticket system, the payment gateways and the hosting — the whole thing. A $49 announcements widget is asking you to double your entire software bill for a bell.

And the free tier puts someone else’s name on your panel

The free plan carries a watermark. It also limits you to one teammate, with no appearance customisation, no scheduling, no pinned posts, no email notifications and zero API requests — but the watermark is the one that matters, because this business has exactly one rule about third parties and it is that the customer never learns there are any. We do not name a provider in a ticket reply. Then a widget in the corner of the same page introduces a different company by name and links to it.

Every panel feed we could find running on that tool was showing the watermark, and by the vendor’s own pricing table the watermark exists only on the free plan. That is an inference rather than something we asked them — but it is a safe one, and it means the thing being sold at $49 is competing with what these panels currently pay, which is nothing.

Check the one you already have first

Before buying anything, ours included: Perfect Panel ships an Updates module of its own that auto-publishes records about service changes, plus separate user notifications, plus one-click integration slots for both Beamer and AnnounceKit sitting in the same list as Tawk.to and Crisp. On 31 July 2026 they added an Updates page to their mobile app, so it is actively being worked on rather than left in place. If the built-in one does what you need, it is free and it is already installed. Go and open it before you read a pricing page.

We are saying that on a page written by a vendor, because the argument does not work unless you have checked. It is also worth knowing that there is no marketplace and no plugin system on that platform — the integration list is hardcoded and first-party — so no third-party tool here is endorsed, vetted or listed by anyone. Every one of them, ours included, has to stand up on its own.

What we built, and why we are not selling it yet

We built one too. It is called Broadcast: a bell and a feed you embed with one script tag, with no visitor meter and no watermark on any plan. It is not on our pricing page, it is not on a plan, and you cannot buy it today — so treat everything in this paragraph as disclosure rather than as an offer.

The reason is the argument at the top of this page. The whole case against the category is that publishing is a second place to type a fact the panel already holds — and the part of Broadcast that reads your service list and posts the change itself is the part that has never run. It needs a panel key it can read, and on the one panel connected to our platform that key cannot be decrypted. So the version that exists does the same job as everything in the table above: you type the post. Selling that beside a product that works, while quoting an argument about retyping, would be the trick this whole page is written against.

What we do sell today is one thing: Pilot Chat, live chat on your panel site that answers a customer out of that panel’s own order records. Broadcast comes back to the pricing page on the day the watcher produces a real catalogue snapshot from a working key, and not before.

Whichever you end up on, the test is the one from the top of this page. Open the feed. Look at the date.