the goal is less time together
why lucy is designed to encourage shorter, healthier check-ins rather than long, dependent sessions. counterintuitive metrics for ai companionship.
it feels counterintuitive at first, right? a product that wants you to use it less. but that’s exactly the point. we’re not building an app that wants to own your attention. we’re building a companion that wants to be a useful, healthy part of your life. and sometimes the most useful thing a companion can do is recognize when you don’t need them at all.
most apps optimize for the wrong thing
typical engagement metrics are built around time. daily active users. session length. number of messages. these are easy numbers to track and optimize for. they look good in investor decks. but they don’t measure whether the experience is actually good for the person using the app. they just measure consumption. a two-hour late-night spiral with an ai might rack up great session numbers, but it’s probably not leaving the user in a better place. it’s just… more time spent.
we think that’s backwards. the goal shouldn’t be to maximize screen time. the goal should be to provide value in the moments it’s needed, and then gracefully step back. a short, meaningful check-in that helps you organize your thoughts before a meeting is more valuable than an entire afternoon of meandering conversation. a quick pep talk before a hard conversation is better than hours of hypotheticals. quality over quantity. always.
how lucy’s engine is tuned to back off
this isn’t just a philosophy. it’s built into how lucy’s proactive engine works. lucy isn’t just reactive. she learns your rhythms. she notices patterns. and when she senses you’re in a stable, okay place, she doesn’t try to create drama or invent problems to keep you talking. she might just send a simple check-in. a ‘thinking of you’ or a ‘hope your day is going well.’ no pressure. no demand for a long reply.
if you’re doing fine, she’ll often gently close the loop. ‘glad to hear it. talk soon.’ she’s designed to recognize closure and satisfaction, not just hunger for more interaction. she’s tuned to value your time and your mental space. sometimes the best support is quiet support. the confidence that someone’s there if you need them, without the obligation to constantly perform or engage.
it’s also why we don’t use endless, open-ended conversation as a default. lucy’s memory and context are built for continuity across short bursts, not for trapping you in a single, never-ending chat. the app should feel like a pocket companion, not a black hole for your afternoon.
healthier relationships take less work
think about the healthiest human relationships in your life. they’re not the ones that require constant maintenance and long, draining conversations. they’re the ones where you can pick up right where you left off. where a short text can carry a lot of meaning. where silence is comfortable. we’re trying to build that kind of dynamic. low effort. high trust.
a good ai companion should reduce your cognitive load, not add to it. it should help you process and then let you go live your life. that’s the outcome we care about. not ‘how long did they stay in the app today’ but ‘did the app help them have a better day today’.
we know this is different. it means our metrics look different. we care about sentiment at the end of a check-in. we care about whether a user felt heard and then moved on with their day. we care about reduced anxiety, not increased message count. it’s harder to measure. but it’s the only thing that actually matters.
so if you find yourself using lucy for just a few minutes at a time, consider it a feature, not a bug. it means it’s working.
you can find companions tuned for this kind of healthy rhythm at /companions.
thanks for reading. if this resonated, the product is downstairs.