| name | optimizing-for-user-feelings |
| description | A product development framework that prioritizes emotional outcomes over quantitative metrics to build high-loyalty consumer products. Use this when defining a new feature's UX, when data-driven optimization has reached a local maximum, or when building products in a commoditized market where brand and delight are the primary differentiators. |
Optimizing for User Feelings
Traditional product development often obsesses over "graphs and metrics" (e.g., clicks, conversion, time in app), which can lead to sterile, uninspired products. This framework shifts the focus to the human on the other side of the screen, identifying the specific emotional state you want to evoke—such as joy, speed, focus, or airiness—and using that as the North Star for design and engineering.
Core Principles
1. Identify the Target Emotion
Before writing a PRD or a line of code, define exactly how the user should feel during and after the interaction. Common target emotions include:
- Lightness/Airiness: For tasks that should feel effortless and non-burdening.
- Surprise/Joy: For features intended to drive organic word-of-mouth.
- Smartness/Focus: For productivity tools that should make the user feel like a "power user."
- Security/Ease: For enterprise or sensitive data interactions.
2. Connect Feelings to Growth
View "Feelings" as a precursor to "Metrics." If a user feels "Surprise" or "Joy," the quantitative result is often viral growth (screenshots shared in Slack, tweets, recommendations). If they feel "Agile," the result is high retention (D5/D7—using the app 5 out of 7 days).
3. Use Metrics as "Honesty Checks"
Metrics should not be the source of inspiration at the moment of creation, but they are essential tools to keep you honest.
- The D5/D7 Metric: Measure how many people use the product at least 5 days a week. This captures retention, engagement, and growth in a single, un-gameable number.
- Cohort Improvement: Don't obsess over absolute numbers; focus on whether the "feeling" you designed for is resulting in better retention curves for newer cohorts.
Application Guide
The "What Could Be" Prompt
Instead of asking "What is the MVP?" or "How do we fix the bug?", ask: "What could be?"
This pushes the team to imagine the most ambitious, "soulful" version of a solution that transcends basic utility.
Resetting Names to Remove Baggage
When building a feature, give it a "made-up" or non-traditional name to shed preconceived notions.
- Bad: "Browser History" (People have a fixed idea of what this looks like).
- Good: "The Archive" or "The Library" (Forces the team to think from first principles about the feeling of looking back).
Swarming for Heartfelt Intensity
Once the target feeling and "what could be" vision are set, "swarm" the project. Hire or assign people with "heartfelt intensity"—those who have something to prove regarding that specific craft (e.g., a performance engineer who wants to achieve double the speed with 25% of the headcount).
Examples
Example 1: The "Peek" Feature (Arc Browser)
- Context: Users on sites like Hacker News want to check 5-10 links quickly without losing their place.
- Target Feeling: "Lightness" and "Airiness."
- Application: Instead of opening a new tab (which feels heavy/permanent), create a temporary overlay that "whoops" up and back down effortlessly.
- Output: The user feels they are "peeking" rather than "browsing," reducing the cognitive load of tab management.
Example 2: Enterprise Software Implementation
- Context: An IT admin setting up a security suite.
- Target Feeling: "Smart" and "Ahead of the curve."
- Application: Design the setup flow to provide immediate, sophisticated insights about their network that they didn't have before, rather than just showing a "Success" checkbox.
- Output: The admin feels empowered and competent, increasing the likelihood they will champion the product internally.
Common Pitfalls
- Romanticizing without Accountability: Optimizing for feelings does not mean ignoring the business. Use a single "hard" metric (like D5/D7) to ensure the feelings are actually translating into usage.
- Choosing Too Many Feelings: Trying to make a feature feel "Fast," "Secure," "Social," and "Fun" all at once leads to a muddled UX. Pick one primary feeling.
- Hiring Only "Specialists": If you only hire people who have done the same role for 10 years, they bring baggage. Hire "mutts"—multidisciplinary people who prioritize "making the thing feel right" over following a standard PM/Designer process.
- CEO-Hero Worship: Don't let the "vision" accrue only to the leader. Publicly celebrate the specific individuals (engineers, designers) who crafted the "feeling" to maintain team intensity.
