Heard
LISTENING REPORT
Prepared for AlexOctober 14, 2026
This was a supportive working conversation in which Alex helped Dana think through a drop in weekday business at her cafe and a looming lease decision.
This report is an assessment, not a grade. It shows how you listened in one real conversation, measured against reference points from counseling research (expert-set, not yet validated for work conversations), so you can choose what to practice next.
The Listening Score
87 out of 100
Counted on the recording.
Room, the share of the conversation you gave them60.0 of 60
Curiosity, how much your questions invited26.9 of 40
Built only from the software counts: the room you gave and the curiosity you showed. It cannot see warmth; the full report below can.
By the numbers, the counted measures
How the airtime and questions actually went
Share of the talking35% of the words, less than half
You spoke about a third of the words and left most of the room to Dana.
Open questions47% of questions, below beginning proficiency
Just under half your questions were open, enough to move Dana from facts to meaning though a few more would deepen it.
Reflections for every question0.47 for each question, below fair
Your reflections were outnumbered by your questions, so you leaned on asking more than mirroring.
Support responses for every shift to yourself12.0 for each shift, more support than shifts
You kept the focus on Dana almost the whole time, with a single shift to your own story.
Coaching signal, a coach's structured judgment
The judgment calls, shown as soft bands
These are not exact scores. They are one structured read of the conversation, made by AI and anchored to published scales coaches and researchers use, adapted to rest on what was said. AI can be wrong, so treat them as a starting point for reflection.
Complex reflections86% of reflections, good
Most of your reflections added meaning Dana had not stated, and she confirmed them in her very next words.
Empathy4.0 out of 5, strong
You tracked what Dana said accurately and reached just past it in ways she took up, landing a solid 4.
Partnership4.0 out of 5, strong
You let Dana's ideas set the direction and checked consent before proposing more, a strong 4.
Our signature measure
Discomfort tolerance
Staying with the hard moment3.5 out of 5, comes and goes
You stayed with the scary parts most of the time but once moved to a solution right after a hard admission.
A strength to keep
You listen your way toward meaning, not just facts. When Dana said the old regulars still say they miss her, you reflected "You missed them, too. It sounds like not just the sales," and she answered "I do. It's a little embarrassing to say I liked being the place where people started their day." You also repair cleanly: after an early suggestion you said "Right. Sorry. I jumped ahead," and handed the thread back to her.
A growth edge
Twice you reached for a fix before Dana asked for one. The clearest was "My uncle had a sandwich shop that went through something like this. He ended up closing on Tuesdays and Wednesdays and his costs went way down." Dana engaged it, but it pulled the focus onto your story right after she had named something scary.
An experiment to try
Next time you feel a solution rising, try parking it behind one more question about their experience, something like "What have you already tried, or thought about trying?" before you offer anything of your own.
Your connection profile
Spaciousness: Made real room, you held 35% of the words and let Dana's ideas lead from the midpoint on.
Curiosity: Genuinely open, your questions moved from story to feeling to "what would make this their place."
Attunement: Accurate and confirmed, reflections like "the group that spends the most is already there" drew "I never thought of it that way."
Steadiness in the hard moment: Mostly steady, you stayed with the March fear but once pivoted to a solution right after it.
The numbers in one line
Talk share 35%; open questions 47% (open/closed is reliable, just under the 50% mark); reflection-to-question ratio 0.47; support-to-shift ratio 12.0; unsolicited suggestions 2.
You tracked Dana's own words and gave them back to her sharpened, and she answered "That's better than I could have said it," then named her own next step: "I need to know who my Tuesday is."
An honest note on what you are reading. The airtime measures (your share of the talking, turn count, and longest turn) are counted by software. The question counts are computed by software from the grammar of each sentence, so they do not vary, though a misheard word in the transcript can change a count. The reflection, advice, and support-to-shift counts are careful structured reads of the transcript, made by AI across several independent passes and merged: close, but not exact to the single instance, and subject to transcript quality. Questions and reflections are counted one per distinct statement, even when several sit inside one speaking turn, which is a finer grain than the research convention and is applied identically to both sides of every ratio. The simple-to-complex split and the 1 to 5 ratings are a coach's structured judgment, made by AI and shown as soft bands. AI can be wrong, so treat its reads as a starting point, not a verdict. This is a communication assessment for capable adults. It is not therapy and never a diagnosis.
How this is scored
What each number measures
- Listening Score: Built from two counts: room given to the other person, up to 60, and open questions, up to 40.
- Share of the talking: How much of the talking the listener did. The report marks 50% as even split.
- Open questions: Questions that ask for more than a yes or no. The coding manual marks 50% as beginning proficiency.
- Reflections for every question: Times the listener said back what they heard, for each question asked. The coding manual marks 1.0 as fair and 2.0 as good.
- Support responses for every shift to yourself: Replies that kept the focus on the other person, for each one that turned it to the listener. The report marks 1.0 as even.
- Complex reflections: Reflections that added a meaning the other person had not said. The coding manual marks 40% as fair and 50% as good.
- Empathy: How well the listener showed they understood, from 1 to 5. The report marks 3.5 as developing and 4.0 as strong.
- Partnership: How much the listener let the other person steer, from 1 to 5. The report marks 3.5 as developing and 4.0 as strong.
- Staying with the hard moment: How well the listener stayed with a hard moment, from 1 to 5. The report marks 3.0 as comes and goes and 4.0 as steady.
Fair and good, on reflections for every question and complex reflections, are the marks of the MITI 4.2.1 coding manual (Moyers, Manuel and Ernst, 2014); beginning proficiency, on open questions, is the MITI 3.1.1 manual's (Moyers and others, 2010). Both manuals say their marks rest on expert opinion; Heard's counts are its own, not a trained coder's.
Airtime and question counts are counted by software the same way every time, and are only as accurate as the transcript. Reflection and ratio measures are tight, moving only a few percent between scorings. The 1 to 5 scales are the median of several independent passes; the looser judgment measures carry real scatter, so read them as ranges: support responses run about 9 to 15, unsolicited suggestions run about 1 to 3.
Where these benchmarks come from. The developing and strong markers on the empathy and partnership scales are reference points from counseling research (expert-set, not yet validated for work conversations), drawn from the motivational interviewing tradition of studying helping conversations. They mark basic competence and proficiency, and the researchers who set them note that they rest on expert judgment rather than large outcome studies, so we treat them as directional guidance. Support and shift responses follow sociologist Charles Derber's research on conversational attention (The Pursuit of Attention, Oxford University Press). Discomfort Tolerance and the Connection Profile are our own coaching lens, presented as coaching signal, not validated scores.