The comment moderation queue we almost didn't build

Six months into Feedr's life, a conference organiser in Manchester messaged us at 11 PM on a Thursday. Her keynote speaker had just finished a talk on a contentious topic. Within ninety seconds, the comment stream filled with seventeen messages. Three were off-topic jokes. Two were hostile. Eight were thoughtful but arrived faster than she could read them. One was a genuine, important question buried at the bottom. She had thirty seconds before the speaker left the stage. She asked us: 'Why can't I just hide the rubbish before it reaches the audience?'

The problem we'd been ignoring

We'd spent months obsessing over the right way to display comments. Should they scroll up or down? How big should the avatars be? What happens when you get a thousand people in a room and they all start typing at once? But we'd missed something obvious: not every comment should go live immediately.

When Feedr launched, we gave hosts a single power: the ability to pin comments they liked. Everything else was automatic. Comment lands in the stream, audience sees it instantly. It worked brilliantly for small groups, casual settings, friendly conversations. For a church pastor taking questions from forty parishioners? Perfect. For a conference speaker standing on stage in front of three hundred people, some of whom were there to test their boundaries? Terrifying.

I realised we'd built Feedr for the ideal audience, not the real one.

The conversation that changed direction

That Manchester message sparked something. I asked our support channel: 'Has anyone else wanted to block or hide comments before they go live?' The replies came in for two days straight. A podcast host. A university lecturer. A pastor whose comments turned into a theological argument within minutes. A corporate event organiser who needed to make sure no one posted competitor brand names. None of them wanted to kill the feature entirely. They just wanted a moment to breathe.

What struck me wasn't the volume of requests. It was the specificity. Each person had a different reason, but they all needed the same thing: control without friction. Not a seven-step approval process. Not an AI system making calls on their behalf. Just a queue. A way to glance at what came in, decide in half a second whether it belonged, and move on.

Building something that had to feel invisible

Here's where it got tricky. If we added moderation, we couldn't make it complicated. A host running a live event has about two seconds of attention to spare. If checking a comment queue felt like opening an email inbox, they wouldn't use it.

We spent weeks talking about what 'invisible' meant. Invisible didn't mean we hid the feature. It meant the feature disappeared into the moment. A small badge on the comment icon. A notification that told you how many messages were waiting. A swipe or tap that took you to the queue, not a separate screen that demanded your focus.

The actual decision mechanism had to be two buttons: approve or hide. No 'mark as spam'. No 'report to admins'. Just does it go live or doesn't it. We also built guest moderators into the same space. If a host needed help, they could send an invite link to someone else on their team, and that person got the same queue, the same two buttons, the same speed.

What we learned from real events

When we shipped the moderation queue to Creator tier users, something unexpected happened. It wasn't that hosts became heavy-handed censors. It was the opposite. Knowing they had a pause button made them braver. One university lecturer told us she actually showed more comments after getting moderation, not fewer, because she felt able to let through the slightly weird questions without worrying about chaos.

The guest moderator feature revealed something too. Hosts weren't using it for censorship. They were using it for speed. A speaker at a three-day conference told us she'd invited her colleague to approve comments while she was talking, so nothing got delayed by more than a second or two. The colleague just sat next to her and tapped through. They made faster, better decisions together than either could alone.

We also discovered that most hosts barely use the hide button. They approve most things. The queue's real job isn't to keep people out. It's to buy thirty seconds of context. A comment that lands while the speaker is mid-sentence might be confusing. The same comment, approved after they've made their point, lands perfectly.

Why we put it behind a tier wall

Some people ask why moderation isn't free. The answer is about intent. Feedr's free tier is for testing, for small groups, for people who just want to see if real-time audience comments work for their format. Most of them don't have a moderation problem because they have twenty people in a room and they know them.

The moment you start running proper events, the moment you care enough to manage your audience's experience with real intention, you shift into Creator. That's the tier that says: 'I'm serious about this. I want unlimited sessions. I want control. I want to own the experience.' Moderation lives there, alongside emoji reactions and comment pinning, because it's part of that ownership.

The feature nobody asked for, but everyone needed

Looking back at that Manchester message, I'm struck by how almost we missed this. The signal was there. We just weren't listening properly. We'd built the first 80% of what hosts needed, then got comfortable. The moderation queue was the last 20%, and it changed everything about how people used Feedr.

The funny part is, we didn't invent moderation. Every successful platform has some version of it. We just did something smaller and more honest than most. We didn't build a 'trust and safety' empire or a content policy machine. We gave hosts a queue, two buttons, and thirty seconds of control. That's all they actually needed.

Have you ever run a live event where the audience moved faster than you could? What would have made the difference?

Want to try Feedr?

Visit Feedr →