two weeks of beta feedback: what we built because you asked

by Wistkey · 8 august 2026 · 5 min read

when we opened the AceSpeak beta, we had a reasonably clear idea of what it should do: run a take, score your gestures, facial expression, pace, intonation and pauses, and give you one clear thing to change next. what we didn't know was what practising that, day after day, would actually feel like. that part, we needed you to tell us.

two weeks in, you have. here's what came directly out of that feedback.

battle mode: challenge a friend, not just yourself

the single most common thing beta testers told us was some version of "this would be more fun with someone else." practising alone works, but it's easy to stop. so we built a head-to-head battle mode: two people run the same prompt, and AceSpeak compares your delivery side by side across all five dimensions, with a radar chart showing exactly where each of you was stronger, a declared winner, and a running tally so you can settle the score across as many rematches as you like.

it turns a slightly awkward solo habit into something you'd actually choose to do with a friend on a tuesday evening. that was the whole point.

a profile that shows your progress, not just your last score

the old flow ended at your latest number. testers wanted to see the shape of their progress, not just the most recent data point: how far they'd come, what tier they'd reached, how consistently they'd shown up. so your profile now carries your rank, your practice streak, your best score, and your badge count, alongside a full tier ladder that runs from silent snail all the way up to galactic ace for the people who really commit to the habit.

it's a small change with a real effect: a badge you're proud of is a much better reason to open AceSpeak tomorrow than a number you half-remember from tuesday.

30-day insights: what's working, and one clear thing to change

this is the one we're most pleased with, because it's the closest to what AceSpeak was always meant to be. every 30 days, AceSpeak now looks back across your practice and tells you, in plain language, what's genuinely working (not generic praise — specific, observable strengths) and one clear thing to focus on next.

that's the same "one-fix" philosophy we started with, just applied over a longer horizon: see it, name it, change one thing, and go again. it turns out testers wanted that logic applied to their whole month, not just their last take.

we're still building

none of these three features were on the roadmap in this form two weeks ago. they exist because people ran takes, opened battles, checked their profile, and then told us honestly what was missing. that's the actual value of running a beta rather than shipping quietly and hoping: AceSpeak gets built with the people who are going to use it, not just for them.

if you're already in the beta, thank you — genuinely, keep the ideas coming. if you're not yet, there's still time to join before it closes, and there's a decent chance your next suggestion ends up in AceSpeak too.