The person nobody was listening to

The problem nobody talks about

When we launched Feedr, I thought the hard part was solved. Get people to scan a QR code instead of downloading an app. Let them comment and upvote in real time. Done. The speaker gets feedback. The audience feels heard. Everyone wins.

Except that's not what happened in practice.

Naomi's comment stream looked like this: genuine question, joke, another joke, someone asking how to get VC funding, actual question, three thumbs-up emoji in a row, someone advertising their podcast, a real insight that took two sentences to explain. All happening in sixty seconds. The speaker literally couldn't parse it. They'd glance at the stream, see chaos, and stop looking.

What Naomi really needed wasn't more features. She needed a filter. A human decision point before comments hit the live feed.

Why we almost missed this entirely

We spent the first six months focused on the wrong metric: volume. How many people can join? How many comments per minute can the system handle? We shipped upvoting, which helped important comments rise to the top. That worked for some people.

But for Naomi, and for dozens of people like her, upvoting wasn't enough. A comment from someone with an agenda, or a drive-by joke, could still muscle its way into visibility just by being first or by landing at the right moment when half the audience was paying attention. And once it was live, it was live. The moment was gone.

I remember reading her message three times. Then I called her. Turned out she'd been running in-person events for eight years. She knew her crowd. She knew which comments would land with the speaker and which would derail them. She just needed the tooling to enforce that judgment call in real time.

Building the queue: the bit that changed everything

We shipped the comment moderation queue on the Creator plan. Here's how simple it is: comments come in, they sit in a queue that only the host sees, and the host taps to approve or skip. The approved comments appear on the live feed. The rejected ones vanish. The audience never knows they submitted something that didn't make the cut. No shame, no friction, no hurt feelings.

What I didn't expect was how much it changed the nature of the conversation itself.

Within a month, Naomi sent another message. This time she said the questions were better. More thoughtful. Longer. People were actually drafting comments instead of just shouting into the void. The upvote counts were lower but the quality was higher. And the speaker was actually reading the feed again.

That part mattered more than the moderation itself. The queue gave the host permission to be intentional. To say 'this comment deserves to be seen by the speaker' instead of 'this comment is loudest'. It's a small difference in tooling. It's a massive difference in outcomes.

The ripple effect nobody planned for

Once we had the moderation queue working, other hosts started using it in ways we hadn't anticipated. One church pastor used it to filter out comments unrelated to the sermon, not to censor but to keep the discussion coherent. A university lecturer used it to catch genuinely confused questions that needed a follow-up explanation, which she'd give in real time. A conference organiser used it to help their panelists by surfacing the most substantive audience questions and burying the ones that were actually just pitches in disguise.

Each of them had a different definition of 'good comment'. But they all had one in common: they wanted to be gatekeepers in a thoughtful way, not just bystanders watching their event get steamrolled.

That's when I realised the moderation queue wasn't a feature. It was permission. Permission to run your event the way you actually wanted to, not the way the platform nudged you to.

What changed about who gets heard

Here's the thing: before the moderation queue, the loudest voices won. Or the fastest. Or the ones who got lucky with timing. It was a meritocracy of attention span and reflexes, not substance.

After the queue, something shifted. The host became the editor. And most hosts turned out to be pretty good editors. They knew what mattered to their speaker. They knew what would land with their audience. They'd spent months or years thinking about this event in a way the algorithm never could.

Naomi told me that since she started using the queue, her attendance has gone up. People stay longer. They ask better questions in the Q&A. Some of them have said the events feel 'less chaotic'. What they mean is: they feel heard, not drowned out.

That's the part that moves the needle. Not the technology. The intention behind it.