The product makes no claims about itself. It is built on behaviors that have been studied for a long time. Here is what those studies found, and how each one shows up in the product. Links go to the primary sources.
Active listening is not a settled matter. In the study of newlyweds by John Gottman and colleagues, paraphrasing a partner's complaint did not predict which marriages lasted, and couples rarely managed it in the middle of a conflict. Some researchers read that as a mark against reflective listening; others read it as a mark against expecting it live, mid-argument.
That is why the product is written and asynchronous rather than spoken, and why reflection is offered as one listening move among several, never required. The listener has time, and a choice.
Sybil Carrère and John Gottman · Family Process, 1999
What it found. How a conflict conversation began predicted how the marriage fared years later; a harsh start rarely recovered.
How the product uses it. The coach helps the first words of a share come out from the writer's side, before anyone answers.
John Gottman, James Coan, Sybil Carrère and Catherine Swanson · Journal of Marriage and the Family, 1998
What it found. Contempt and defensiveness, and a failure to de-escalate, distinguished the couples who later divorced; paraphrasing did not predict outcomes and was rarely done.
How the product uses it. Contempt is coached or blocked. De-escalation is built in as the pause. Reflection is offered, not required.
John Gottman and Lowell Krokoff · Journal of Consulting and Clinical Psychology, 1989
What it found. Anger and open disagreement did not predict decline in satisfaction over time; defensiveness, stubbornness and withdrawal did.
How the product uses it. Anger and hard topics are allowed in a share. Only contempt and attacks are not.
John Gottman and Robert Levenson · Journal of Personality and Social Psychology, 1992
What it found. Physiological arousal during conflict, what Gottman calls flooding, predicted later dissolution; a flooded body needs time, on the order of 20 minutes, to settle.
How the product uses it. A pause must be at least 20 minutes and has a return time.
Harry Reis and Shelly Gable · Current Opinion in Psychology, 2015
What it found. Feeling understood, validated and cared for by a partner, perceived partner responsiveness, sits at the center of intimacy and closeness.
How the product uses it. The only success mark in the product is the sharer saying they felt understood. Nothing else counts.
Shelly Gable, Harry Reis, Emily Impett and Evan Asher · Journal of Personality and Social Psychology, 2004
What it found. Sharing good news with a partner, and getting an active, interested response, was linked with wellbeing and relationship quality.
How the product uses it. "Something good" is a first-class kind of share, with its own listening moves: celebrate, ask to hear more, extend.
Shelly Gable, Gian Gonzaga and Amy Strachman · Journal of Personality and Social Psychology, 2006
What it found. How partners responded to each other's good news predicted relationship wellbeing and stability at least as well as how they handled conflict.
How the product uses it. The coach catches deflating responses to good news, the "that's nice, but…", and offers an engaged one.
Howard Markman, Mari Jo Renick, Frank Floyd, Scott Stanley and Mari Clements · Journal of Consulting and Clinical Psychology, 1993
What it found. Couples trained in a structured speaker–listener format, one has the floor and the other paraphrases without rebuttal, kept higher communication quality years later than untrained couples.
How the product uses it. The floor, the listening moves, and the rule that disagreement gets its own turn come from this structure.