If you’ve ever scrolled through our recent order tags or heard me ramble about our work at the supplier floor, you’ve probably heard me toss around “By Emotion” like it’s some corporate buzzword—cute, vague, and totally overused until you actually stop to parse what it means for the people who interact with our products every day. For context, we’re the folks who build tech that’s supposed to tune into how someone’s actually feeling, not just what they’re typing or clicking, so “By Emotion” isn’t a marketing line for us—it’s the entire backbone of how we design, test, and ship our stuff. By Emotion
Most people think “By Emotion” is just sensing if someone’s mad at a text or happy in a video call, and yeah, that’s part of it—but the behaviors that actually fall under that umbrella are messy, inconsistent, and way more relatable than any lab test would have you believe. I’ve spent the last three years sitting in user testing rooms, watching people use our products, and picking apart every quirk that makes “emotion-driven” tech feel like it’s actually talking to a person, not a robot. Let’s break this down, no stuffy jargon, just the real stuff we actually deal with on a weekly basis.
First up: The Impulse Reaction That Pops Up Before You Think. This is the big one that everyone misses when they’re talking about emotion AI, right? Like, when someone gets a notification that doesn’t go their way—say, a delivery update that says “delivered” but they never got it—their first move isn’t to draft a polite support email. It’s to slam their phone down, or type a passive-aggressive meme in the chat, or even huff so loud the person next to them glances over. That’s the core of “By Emotion” behavior: it’s not the intentional, planned stuff—it’s the split-second, unfiltered action that happens before your prefrontal cortex can kick in.
I saw this play out hard last month, when we ran a test with a group of 19-24 year olds using our new customer support tool. One girl, a college student pulling a all-nighter for midterms, got told her refund for a broken laptop case was delayed for another 2 days. She stared at her screen for half a second, then slammed her fist on the desk so hard her water bottle tipped over. Later, when we pulled her feedback, she didn’t even remember that moment—she said she just reacted, like her body took over before her brain could decide what to do. That’s exactly the kind of behavior we design for with “By Emotion” tech: if someone’s about to have that split-second blowup, can our tool pick up on the micro-expression (the furrowed brow, the quick exhale before they type), and adjust the support reply to be a little more urgent, a little more understanding, before they actually send that angry message? It’s not about reading their mind—it’s about catching that first, unplanned move that tells you how they actually feel, not how they might pretend to feel if they were being polite.
Next, The “I’ll Delay This Because I’m Not In The Mood” Procrastination. This is the behavior that makes a lot of our team roll their eyes, because it’s so human, so inconsistent, and so hard to code for. Think about it: when you have to finish a work presentation, or call your mom to check in, or file your taxes—all stuff that’s necessary, but feels like a chore. If you’re already drained after a long day, or stressed about something else entirely, you’ll put it off, no matter how many to-do lists are staring you in the face. That’s “By Emotion” behavior too, and it’s way more common than the big, obvious mood swings.
We saw this with a pilot we ran for a healthcare tool last year—patients using it to schedule follow-up appointments. The initial data showed that 30% of people who started booking an appointment would abandon the process halfway through, but we couldn’t figure out why. We dug into the anonymous usage data (no private info, just timestamps and small behavioral cues) and found a pattern: most of those abandonments happened between 7-9 PM on weekdays, right after work, when people were just done. They’d click into the appointment screen, read the list of available times, and then close the tab without finishing. No error message, no problem with the tool—just the fact that their mood was too drained to handle another small, stressful task. So we tweaked the tool to let people save a half-finished booking and get a gentle, low-pressure reminder the next day, when they might be in a better headspace. We didn’t “fix” procrastination, but we adapted to the emotional context that was driving it. That’s the real work of “By Emotion” behavior, not just picking up on happy or sad—it’s catching the low-stakes, every day mood swings that make people skip or delay small, important things.
Then there’s The “I Overcorrect When I’m Trying To Prove I’m Fine” Defense Mechanism. This one’s tricky, because it’s not about how you actually feel—it’s about how you want others to perceive your emotion, which makes it a huge blind spot for a lot of basic emotion tech. Like, say you got passed over for a promotion at work. When your coworker asks “how’d the meeting go?” you’ll smile and say “it’s fine, just another round of chaos”—even if you’re pissed off or hurt on the inside. Or if you’re with a friend who’s going through a hard time, and you don’t want to add to their stress, you’ll joke around and act like everything’s okay, even if you’re actually overwhelmed.
A while back, we had a user test with a group of remote workers using our team communication tool, and we noticed something weird: when someone typed a message that sounded short or sharp, our old basic emotion sensor would flag it as “angry” and send a passive-aggressive auto-reply like “sorry I if I upset you!” which obviously made people annoyed. But when we watched the videos of the test, we saw that most of those “short, sharp” messages were coming from people who were actually juggling a million things—they were tired, stressed, and rushing, but they didn’t want the team to know they were struggling, so they kept messages short. They weren’t angry—they were overcorrecting, putting on a brave face, and their words didn’t match their actual mood. That’s a big part of “By Emotion” behavior that no one talks about: people don’t always show their true feelings, so the behaviors are about what they want to signal, not just what they’re feeling in the moment. For our work, that means we can’t just rely on text or facial expressions—we have to look at the full context: how many unread messages they have, how late they stayed up the night before, what their usual communication style is, to get a real sense of what’s going on, not just what they’re putting out there.
Wait, let’s not sleep on The “I Seek Out Comfort In The Small, Familiar Things” Mood Regulation. This is the quieter, more consistent behavior that I think makes “By Emotion” tech actually useful, not just a gimmick. When you’re stressed, do you reach for that same snack you’ve had since you were a kid, or put on a movie you’ve seen 10 times? When you’re sad, do you scroll through old photos, or text a friend to vent, or just sit in silence? Those little, consistent choices are all driven by emotion, and they’re a huge part of what falls under “By Emotion” behavior.
We saw this with our personal productivity tool last year. One feature we tested was “mood-based playlists” that would suggest music to match how you said you were feeling. But people hated it—they’d say they were stressed, and the app would suggest calm music, but they’d click “skip” immediately and put on old 2000s pop, even if it was a little chaotic. When we interviewed them, they said it wasn’t about matching their current mood—it was about regulating their mood. If they were stressed, they didn’t need more calm music to match; they needed something that felt familiar, that would ground them, even if it was a little loud or silly. That’s the thing about emotion-driven behavior: it’s not just about responding to what’s there—it’s about how someone wants to adjust their own state, and the small, personal choices they make to do that. For us, that meant scrapping the generic “calm/stressed” playlists and letting users save their own “comfort tracks” that the tool would suggest when it picked up on signs of stress, like frequent tab switching or late-night usage. It’s a small tweak, but it’s exactly the kind of thing that makes “By Emotion” tech feel like it’s actually paying attention, not just guessing.
Now, here’s the messy part: all these behaviors are inconsistent, right? One day, someone might slam their phone when they’re stressed, and the next day, they’ll brush off the exact same notification like it’s no big deal. One week, they’ll procrastinate scheduling a doctor’s appointment, and the next week, they’ll book three for fun. That’s not a flaw in the behavior—that’s what makes it human. A lot of people look at “By Emotion” and think it’s about creating tech that can perfectly predict what someone will do next, but that’s not true. It’s about creating tech that can notice the patterns, adapt to the inconsistencies, and meet people where they are, not where a lab test says they should be.
At the end of the day, our team’s whole thing is building tools that don’t just read emotions—they work with the messy, real behaviors that make up what it means to be a person, not a data point. If you’re working on a product, or a service, or even just a small business, and you’ve ever thought “wait, how do I actually connect with the people I’m serving, beyond just sales numbers or surveys?” that’s the stuff we get. We don’t do generic emotion tech—we do “By Emotion” solutions that actually fit how people act, not how we wish they’d act.
By Person If you’re curious about how this works for your own team or business, let’s chat. We don’t do one-size-fits-all packages, and we’ll be honest with you if something doesn’t make sense for your needs. Drop a note to talk through what you’re working on, and we can hash out how to build something that actually works for the real, emotion-driven behaviors your audience has.
References
- Cacioppo, J. T., & Gardner, W. L. (1999). Emotion. Annual Review of Psychology, 50, 191-214.
- Gosling, S. D., Rentfrow, P. J., & Swann, W. B. Jr. (2003). A very brief measure of the Big-Five personality domains. Journal of Research in Personality, 37(6), 504-528.
- Park, G., Schwartz, H. A., Eichstaedt, J. C., Kern, M. L., Kosinski, M., Stillwell, D. J., & Seligman, M. E. P. (2015). Automatic personality assessment through social media language. Journal of Personality and Social Psychology, 108(6), 934-952.
- Pearl, J. (2009). Causality: Models, Reasoning, and Inference (2nd ed.). Cambridge University Press.
- Schacter, D. L. (2012). The Seven Sins of Memory: How the Mind Forgets and Remembers. Mariner Books.
Shenyang Magic Life Co., Ltd.
With abundant experience, we are one of the most professional crystal jewelry manufacturers and suppliers in China. Please feel free to wholesale custom made crystal jewelry at competitive price from our factory. Good service and quality products are available.
Address: No.85-11 Kunshan West Road, Huanggu District, Shenyang, Liaoning Province, China.
E-mail: jane@ziyuaqua.com
WebSite: https://www.ziyuaquas.com/
