The marketing metric nobody tracks (but probably should)
Most product teams track acquisition, activation, retention. The usual funnel. We track all of that at Murror too.
But there's one metric we started paying attention to that changed how we think about growth entirely: how often users talk about themselves differently after using the product.
Not NPS. Not referrals. Not even engagement time. We look at the language shift.
When someone starts a session saying "I don't know why I keep doing this" and ends it saying "I think I do this because..." -- that's the moment we care about. It's not a conversion. It's a transformation.
Here's why this matters for marketing:
Traditional growth metrics tell you what people do. Language shift tells you what people become. And people who feel like they're becoming something don't churn. They don't need to be re-engaged with push notifications or discount emails. They come back because the product changed how they see themselves.
We started noticing this by accident. Our best word-of-mouth didn't come from users saying "Murror is great, you should try it." It came from users saying things like "I realized something about myself" in conversations with friends -- and those friends asking "how?"
The product wasn't the story. The personal shift was.
I think there's something here for anyone building products that go deeper than surface-level utility. If your product actually changes how people think or feel, your best marketing metric might not be in your analytics dashboard at all. It might be in the conversations your users have when you're not in the room.
Anyone else tracking something like this? Or found unconventional signals that predict growth better than the standard metrics?
Replies
Do users become more likely to return after these reflection moments happen?
Murror
@graham_lewis Yes, significantly. We noticed that users who experience at least one clear language shift in their first week have about 3x higher 30-day retention compared to those who don't. It seems like once someone sees themselves differently through the product, it creates a reason to come back that goes beyond habit or notifications.
This really stood out to me. People don't usually tell friends about a feature, they talk about the moment something clicked for them . That kind of shift feels way more meaningful than a lot of traditional growth metrics.
Murror
@james_anderson77 Exactly. The best word-of-mouth doesn't come from feature satisfaction -- it comes from identity shifts. When someone says "I realized something about myself," that's the kind of story people naturally share. We've found that this type of organic referral converts at a much higher rate too, because the person hearing it already understands the emotional value before they even try the product.
If you've found any reliable way to measure that at scale, or is it still more of a qualitative, "you know it when you see it "signa.?
Murror
@jared_coleman Great question. It started qualitative, but we've built it into something more systematic. We use a combination of sentiment analysis and keyword pattern matching to flag potential language shifts, then validate a sample manually each week. It's not perfect -- we still miss some subtle ones and occasionally get false positives -- but it's reliable enough to track trends and compare cohorts. The key was accepting that it doesn't need to be precise to be useful. Even a rough signal of identity shift turned out to be more predictive of retention than our most polished traditional metrics.
What made you first realize language change was worth tracking as a signal?
Murror
@ashton_blake It was actually an accident. We were doing user interviews and kept hearing people describe themselves differently at the end of the call than at the beginning. One person started with "I'm not good at understanding my feelings" and ended with "I think I'm starting to see patterns in how I react." That contrast was so striking that we went back through our session data and started looking for similar shifts. Once we saw how consistently it correlated with long-term engagement, we knew we had found something worth measuring.
Do users openly describe these realizations themselves, or is it something your team inferred over time?
Murror
@aurora_parker Both, actually. Some users are very explicit about it -- they'll write things like "I never thought of it that way before" or "I didn't realize I was being so hard on myself." But we also use NLP to detect subtler shifts, like when someone moves from passive language ("things happen to me") to more agentic language ("I chose to" or "I noticed that I"). The combination of self-reported moments and the patterns we detect gives us a more complete picture.
Curious whether you've noticed certain product moments or prompts consistently triggering these language shifts more than others?
Murror
@rivra_dev Definitely. The moments that trigger the biggest shifts tend to be ones where the product reflects something back that the user hadn't articulated yet. For Murror, it's often when the AI summary of a journal entry names an emotion or pattern the user was feeling but hadn't put into words. There's something powerful about seeing your own experience described clearly by something outside yourself -- it creates a moment of recognition that often leads to deeper self-reflection.
@monatruong_murror
This is honestly one of the most interesting retention discussions I’ve seen in a while.
A lot of products measure whether users return, but not whether the user’s internal narrative changed after interacting with the product.
What stood out to me is the idea that people come back when they feel a shift in identity, not just utility.
I’ve been noticing something similar while building Gleyo around onboarding + retention flows for communities.
Many products optimize for task completion:
joined
clicked
finished onboarding
But the stronger signal is usually behavioral or emotional:
did the user now see themselves as “part of this”
did they form a reason to return without being pushed
did the experience create continuity in their mind
That’s usually where long-term retention starts.
The “language shift” framing is really smart because it captures transformation, not just activity.
Murror
@okiri_donald This is a really great breakdown, and I love how you've framed the distinction between task completion signals and behavioral/emotional ones. Your three questions -- do they see themselves as part of this, do they form a reason to return, does the experience create continuity -- are essentially what we're trying to measure with language shift, just from different angles. For communities especially, I imagine that "part of this" signal is even more powerful because the identity isn't just personal, it's social. Would love to hear more about how you're detecting those signals at Gleyo.
@monatruong_murror Yeah exactly, I think for communities the identity layer becomes very social.
A lot of onboarding systems measure whether users completed steps, but not whether they developed any sense of belonging or continuity afterward.
One thing I’ve been noticing while building Gleyo is that retention tends to strengthen once users stop feeling like “visitors” and start feeling like participants with context, relationships, or momentum inside the ecosystem.
Sometimes that signal appears in subtle behaviors:
returning without prompts, checking updates voluntarily, bringing friends in, or participating even when there’s no direct reward attached.
I’m still exploring how to measure those signals more intentionally, but they seem far more predictive of long-term engagement than surface-level activity metrics alone.
This is really interesting framing. The metric you're describing sounds like a version of what BJ Fogg calls "identity reinforcement" the product isn't just solving a problem, it's telling users something about who they are. The products I've kept using longest are always the ones where the act of using them made me feel like a slightly better version of myself, not just more efficient. Hard to measure but probably the best leading indicator for retention. Curious how you're actually capturing it survey? behavioral signal?
Murror
@sagar_kalra1 Love the BJ Fogg connection -- identity reinforcement is exactly the right frame. To answer your question: it's mostly behavioral signal, but we layer in both. On the behavioral side, we use NLP to detect shifts in how users describe themselves and their experiences across sessions. On the survey side, we do occasional lightweight check-ins that ask users to reflect on how they see themselves now vs. when they started. Neither is perfect alone, but combined they give us a much richer picture than any single retention metric. The behavioral signal is noisier but scales better; the survey signal is cleaner but harder to get consistently. We've found the behavioral side to be more predictive of long-term retention even with the noise.