Continue learning

You have the handbook’s vocabulary and plays. The next step is not to collect more bookmarks—it is to deepen the few questions that still feel thin for your product.

Start from a real gap. Maybe first value still feels muddy, recovery after error feels cold, or the paywall still smells like pressure. Name that gap in the language of this handbook—feeling north star, trust, agency, a strategy or TTP—then go reading with that question in hand. When you come back, ask what changed in your north star or which TTP you would use differently. If nothing changed, the reading was entertainment, not learning.

Where to go, depending on what you need

If you want the science of feeling and craft. Don Norman’s essays at jnd.org and Aarron Walter’s Designing for Emotion are short, durable foundations. Nielsen Norman Group is the workhorse for usability, onboarding, trust, and empty states—search a specific moment (“error messages”, “permission dialogs”) rather than browsing the homepage. The Interaction Design Foundation is useful when you need a clear primer on a mechanism before you argue about taste. For empathy methods and design-thinking prompts, Stanford d.school resources and the IDEO Journal still earn their keep; Google’s Design Library is better for concrete UX case studies than for theory.

If you are still unsure who the customer is. Stay close to discovery practice: Teresa Torres / Product Talk on continuous interviewing, Marty Cagan / SVPG on discovery as product work, and the Y Combinator Library on talking to users without fooling yourself. Pair that with this handbook’s Customer Discovery so the interviews turn into something your team—and your agents—can load.

If the problem is product judgment, not a missing widget. Lenny’s Newsletter, First Round Review, and April Dunford on positioning help when the question is “what are we even trying to be?” Reforge essays are useful for activation and retention loops—read them as hypotheses to test against your feeling north star, not as a growth checklist. a16z Ideas is worth skimming when market shape, platforms, or AI product bets are the fog—not when you need a Tuesday shipping decision.

If behaviour and habit are the terrain. Behavioral Scientist and The Decision Lab keep you honest about motivation and bias. Behavioral Design Academy is explicitly oriented toward ethical application. Treat habit literature carefully: some of it explains return; some of it sells compulsion. Read Nir and Far critically—useful for naming hooks, dangerous if you mistake addiction patterns for product quality. Prefer sources that protect agency. When a technique only works if the user feels trapped, it does not belong here.

If you are building with AI. People feel uncertainty, misplaced confidence, and loss of control around models. The Google People + AI Guidebook and Microsoft’s Human–AI Interaction guidelines are practical; IBM Design for AI is especially useful in enterprise and high-stakes contexts. Lab blogs from Anthropic, OpenAI, and similar are useful for patterns of interaction, not for copying marketing claims. Ask: how does this product communicate doubt, invite correction, and fail without blaming the user?

If the stakes are trust, safety, or harm. High-stakes flows need more than polish. Center for Humane Technology, NIST’s AI Risk Management Framework, and the responsible-AI material from Microsoft and Google are for when your product can scare, mislead, or lock someone in. Read them when the feeling you want is safety—and when the feeling you must never create is helplessness.

If the product is a journey, not a screen. Service design reminds you that people remember arcs: before the app, during the wait, after the email. The Service Design Network and studios that write about end-to-end experience (for example Frog Design Mind) help when your problem lives between surfaces—alongside the IDEO material already noted above.

Steal feeling from products you admire

Pick two or three products that already feel the way you want yours to feel—speed, calm, confidence, clarity—and study sessions, not feature pages. Walk a first-run, an error, a paywall, a return visit. Write down the feeling in one sentence, then name the handbook idea it maps to.

The engineering blogs of Linear, Stripe, PostHog, Vercel, and Notion are useful when they show how craft decisions were made. Ignore the launch announcements; keep the post-mortems and design write-ups.

How to know you are done with a rabbit hole

You are done when you can answer three questions out loud:

1. What did I believe about our product’s feeling that I no longer believe?
2. Which strategy or TTP would I change this week because of that?
3. What evidence would prove me wrong again?

If you cannot answer those, keep reading—or stop and go talk to a customer instead. More sources will not fix an untested north star.

Bring what you learn back into Why it Works for the lineage, Concepts for shared language, and the strategy or TTP you are actually shipping against. Learning that never touches the product is just a nicer form of procrastination.