A Telegram UX frustration taught me a distinction I now think about in everything we design: where information lives is not the same question as where attention goes.
It started as a small complaint.
We had just moved our conversation from WhatsApp to Telegram, and I found myself getting lost. Not lost in the sense of not understanding the ideas — lost in the interface itself. Topics, threads, channels. Each one made sense on its own. Together, I couldn’t find our conversation anymore.
I said as much to the team: I got a bit lost in all the Telegram channels and threads, and it made me realize that, for me, our WhatsApp group was currently the only place where the conversation still felt coherent. I wondered whether we should create one additional Telegram group, just the three of us plus EMET, no topics, just one continuous conversation. EMET would listen and learn from everything there, but never participate or respond. Our human room, while the other Telegram spaces stayed the places where we actively work and interact with EMET.
The answer that wasn’t wrong, but wasn’t enough
Victor’s first instinct was reasonable: that’s what the #general thread is for. The other threads need to be organised with a concrete purpose and process, the way he was already doing with the Ideation group and the way I’d done with the TL;DR and Value groups — a clear connection from theme, to purpose, to process.
Conceptually, I agreed completely. #general should play exactly that role.
But the problem I was experiencing wasn’t conceptual. It was the Telegram UX itself: when there’s activity in other topics, #general moves down the topic list and stops feeling like the “home” of our conversation. The information was all there, correctly organised, correctly labelled. It just wasn’t where my attention landed.
That’s when I said the sentence that’s been sitting with me since: I think the distinction is important — it’s not only information architecture, it’s also attention architecture.
Two different design questions
Information architecture asks: where does this belong? What’s the right category, the right topic, the right thread? Get that right, and in principle everything is findable.
Attention architecture asks a different question: where does a person’s eye actually go, moment to moment, when ten other things are also asking for it? Get information architecture right and get attention architecture wrong, and you still end up with people feeling lost — not because the system failed to store anything, but because it failed to surface the right thing at the right moment.
I wasn’t asking for another conceptual channel. I already had one — #general was functioning exactly as designed. I was asking for our human conversation to remain visible as a first-class chat, not a filing location that happened to also contain us. That’s not a storage problem. It’s a visibility problem, and visibility is a design choice, not a side effect of correct categorisation.
Victor said yes immediately, and we created the group. It turned out to be a small, almost administrative decision — one new Telegram group, no new process. But I think the principle underneath it is bigger than the decision itself, and it’s one I keep running into now that I’m looking for it: a team, and increasingly a human-agent team, needs both kinds of architecture designed on purpose. One for where things are kept. One for where attention actually goes when everything is competing for it.
We got the first one right from the start. The second one, we only noticed by living inside what we’d built and feeling where it didn’t work.
← All posts
