{"id":"47056ebf-1944-486f-b89e-b3d5874d4a8c","shortId":"AjqeHG","kind":"skill","title":"behavioral-product-design","tagline":"Apply behavioral science to product design: target behavior, intervention map, experiment plan.","description":"# Behavioral Product Design\n\n## Scope\n\n**Covers**\n- Turning a desired user behavior into an **executable design + experiment plan**\n- Diagnosing behavior using **barriers/drivers** (motivation, ability/friction, uncertainty, habit, context)\n- Designing **behavioral interventions** (e.g., defaults, commitment devices, loss aversion/progress, reducing uncertainty) with ethical guardrails\n- Producing decision-ready artifacts a PM/Design/Eng team can build and test\n\n**When to use**\n- “Help me apply behavioral science / behavioral economics to this flow.”\n- “We need to improve retention / activation / onboarding completion.”\n- “Design a streak / habit loop / reminder system (without being spammy).”\n- “Users procrastinate (present bias). How do we get them to do the thing?”\n- “People stick with the status quo. How do we drive switching/adoption?”\n- “Users are uncertain / anxious. How do we reduce uncertainty and move them forward?”\n\n**When NOT to use**\n- You need upstream strategy first (vision, positioning, roadmap). Use `defining-product-vision` / `prioritizing-roadmap`.\n- You can’t name the target user + target behavior + success metric (this becomes generic advice).\n- The goal is to create **dark patterns** (deception, coercion, addiction, hidden costs). Don’t do this.\n- The domain is regulated/high-stakes (medical, financial advice, minors). Require domain/legal review and tighter safeguards.\n- You need to design an onboarding flow without a behavioral science lens -> use `user-onboarding`.\n- You need to analyze retention/engagement metrics and cohort data, not design interventions -> use `retention-engagement`.\n- You need to design a survey or research study to collect user data -> use `designing-surveys`.\n- You need to test an existing design with real users -> use `usability-testing`.\n\n## Inputs\n\n**Minimum required**\n- Product context + target user segment\n- The **target behavior** (what user action you want more of, in what context)\n- Baseline funnel/retention metrics (even rough) + where the drop happens\n- Constraints: platform (web/mobile), notification channels, brand/tone, time box\n- Existing evidence: user research notes, support tickets, analytics, session replays (if any)\n\n**Missing-info strategy**\n- Ask up to 5 questions from [references/INTAKE.md](references/INTAKE.md).\n- If answers aren’t available, proceed with explicit assumptions and label unknowns. Offer 2 scopes: **narrow (1 behavior)** vs **broad (journey)**.\n\n## Outputs (deliverables)\n\nProduce a **Behavioral Product Design Pack** (in-chat as Markdown; or as files if requested), in this order:\n\n1) **Context snapshot** (goal, segment, constraints, baseline)\n2) **Target behavior spec** (behavior statement + success metric + guardrails)\n3) **Behavioral diagnosis** (barriers/drivers; where bias/friction/uncertainty shows up)\n4) **Intervention map** (ideas mapped to journey moments + mechanism + risk)\n5) **Prioritized intervention shortlist** (top 1–3 with rationale)\n6) **Behavioral design specs** (1–3 build-ready “intervention cards”)\n7) **Experiment + instrumentation plan** (events, primary/guardrail metrics, rollout/rollback)\n8) **Risks / Open questions / Next steps** (always included)\n\nTemplates: [references/TEMPLATES.md](references/TEMPLATES.md)\n\n## Workflow (8 steps)\n\n### 1) Intake + define the target behavior\n- **Inputs:** User context; [references/INTAKE.md](references/INTAKE.md).\n- **Actions:** Clarify the user, context, and *one* primary target behavior. Define success + guardrails (what must not get worse).\n- **Outputs:** Context snapshot + target behavior spec.\n- **Checks:** Target behavior is observable and time-bounded (not “be more engaged”).\n\n### 2) Map the current journey + “moments that matter”\n- **Inputs:** Current flow/JTBD; baseline funnel.\n- **Actions:** Sketch the steps from trigger → action → outcome. Mark drop-offs and emotional moments (uncertainty, effort, waiting, completion).\n- **Outputs:** Journey map summary + top 3 friction points.\n- **Checks:** Each friction point is tied to a specific step/state (not a vague complaint).\n\n### 3) Run a behavioral diagnosis (barriers + drivers)\n- **Inputs:** Journey moments; evidence; assumptions.\n- **Actions:** For each friction point, identify: (a) motivation/benefit perception, (b) ability/friction, (c) prompts/forgetting, (d) uncertainty/risk perception, (e) social/context constraints. Map likely mechanisms (e.g., present bias, status quo, uncertainty aversion, loss aversion/progress).\n- **Outputs:** Behavioral diagnosis table (barrier → mechanism → design implication).\n- **Checks:** Each proposed mechanism has at least one supporting signal (research/quote/data) or is labeled “hypothesis”.\n\n### 4) Generate intervention ideas (mechanism-first, not UI-first)\n- **Inputs:** Diagnosis table.\n- **Actions:** Brainstorm 2–4 interventions per priority barrier using the pattern library in [references/WORKFLOW.md](references/WORKFLOW.md) (defaults, reducing uncertainty, progress/loss framing, commitment devices, reminders, celebration/pause moments).\n- **Outputs:** Intervention inventory (10–20 ideas) with mechanism tags.\n- **Checks:** At least one idea reduces friction (ability) and one reduces uncertainty (trust), not only “add reminders”.\n\n### 5) Add resilience + reinforcement (without manipulation)\n- **Inputs:** Intervention inventory.\n- **Actions:** For habit/retention loops, explicitly design: (a) **reinforcement** (“pause moments” for meaningful progress), (b) **resilience** (“bend not break” policies like grace periods), (c) ethical framing (user benefit, transparency, easy opt-out).\n- **Outputs:** Updated interventions with reinforcement/resilience + ethics notes.\n- **Checks:** No intervention relies on deception, forced continuity, or hidden penalties.\n\n### 6) Prioritize and pick the top 1–3 bets\n- **Inputs:** Updated inventory; constraints.\n- **Actions:** Score ideas on impact, confidence, effort, and risk (trust/legal/brand). Pick 1–3 that cover different failure modes (friction vs uncertainty vs motivation).\n- **Outputs:** Prioritized shortlist + “why these” rationale.\n- **Checks:** Each selected bet has a clear hypothesis and measurable metric movement.\n\n### 7) Write build-ready behavioral design specs + experiment plan\n- **Inputs:** Shortlist; [references/TEMPLATES.md](references/TEMPLATES.md).\n- **Actions:** For each bet, write an intervention spec: hypothesis, mechanism, UX/copy, states, edge cases, instrumentation, rollout/rollback, and guardrails.\n- **Outputs:** 1–3 behavioral design specs + experiment/instrumentation plan.\n- **Checks:** Engineering can implement without major missing decisions; measurement is feasible.\n\n### 8) Quality gate + finalize\n- **Inputs:** Draft pack.\n- **Actions:** Run [references/CHECKLISTS.md](references/CHECKLISTS.md), score with [references/RUBRIC.md](references/RUBRIC.md), and add **Risks / Open questions / Next steps**.\n- **Outputs:** Final Behavioral Product Design Pack.\n- **Checks:** The pack is specific to this product and can be executed in 1–2 sprints.\n\n## Quality gate (required)\n- Use [references/CHECKLISTS.md](references/CHECKLISTS.md) and [references/RUBRIC.md](references/RUBRIC.md).\n- Always include: **Risks**, **Open questions**, **Next steps**.\n\n## Examples\n\n**Example 1 (Activation):** “New users abandon setup on step 3. Use behavioral science to redesign onboarding and propose 2 experiments.”  \nExpected: diagnosis of the abandonment moment, intervention map, 2 intervention specs, and an experiment + instrumentation plan.\n\n**Example 2 (Retention/habit):** “We want a 7-day habit loop for daily check-ins without annoying notifications.”  \nExpected: habit/reinforcement plan (incl. bend-not-break), celebration moments, a streak spec, and guardrail metrics.\n\n**Boundary example (redirect):** “We need to analyze our retention cohorts and understand where users are churning.”\nResponse: redirect to `retention-engagement` -- this request needs metric analysis and cohort diagnostics, not behavioral intervention design. Come back to behavioral-product-design once you know *where* and *why* users drop off.\n\n**Boundary example (ethical refusal):** “Make the UI more addictive so people can’t stop using it.”\nResponse: refuse dark patterns; reframe toward user-beneficial behaviors, transparency, and opt-out controls.\n\n## Anti-patterns\n\nAvoid these common failure modes when applying behavioral science to product design:\n\n1. **Bias-name-dropping without diagnosis** -- Listing cognitive biases (anchoring, loss aversion, social proof) without mapping them to specific friction points in the user journey. Every cited bias must connect to a concrete step where users drop off or hesitate.\n2. **Notification-as-intervention** -- Defaulting to push notifications and reminders as the primary behavior change tool. Notifications address forgetting but not motivation, ability, or uncertainty. Cover all barrier types.\n3. **Dark pattern disguised as nudge** -- Using behavioral techniques to trick users (hidden costs, forced continuity, confirm-shaming). Every intervention must pass the transparency test: would the user agree this helps them if you explained it?\n4. **Generic habit loop** -- Applying a cookie-cutter trigger-action-reward loop without diagnosing the specific barriers for this user segment. Habit design must be grounded in the actual journey data and friction points.\n5. **Missing guardrail metrics** -- Designing interventions to increase a target behavior without tracking unintended side effects (e.g., increased task completion but lower satisfaction, or higher engagement but more support tickets).","tags":["behavioral","product","design","lenny","skills","plus","liqiongyu","agent-skills","ai-agents","automation","claude","codex"],"capabilities":["skill","source-liqiongyu","skill-behavioral-product-design","topic-agent-skills","topic-ai-agents","topic-automation","topic-claude","topic-codex","topic-prompt-engineering","topic-refoundai","topic-skillpack"],"categories":["lenny_skills_plus"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/liqiongyu/lenny_skills_plus/behavioral-product-design","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add liqiongyu/lenny_skills_plus","source_repo":"https://github.com/liqiongyu/lenny_skills_plus","install_from":"skills.sh"}},"qualityScore":"0.474","qualityRationale":"deterministic score 0.47 from registry signals: · indexed on github topic:agent-skills · 49 github stars · SKILL.md body (9,645 chars)","verified":false,"liveness":"unknown","lastLivenessCheck":null,"agentReviews":{"count":0,"score_avg":null,"cost_usd_avg":null,"success_rate":null,"latency_p50_ms":null,"narrative_summary":null,"summary_updated_at":null},"enrichmentModel":"deterministic:skill-github:v1","enrichmentVersion":1,"enrichedAt":"2026-04-22T06:56:18.784Z","embedding":null,"createdAt":"2026-04-18T22:16:14.512Z","updatedAt":"2026-04-22T06:56:18.784Z","lastSeenAt":"2026-04-22T06:56:18.784Z","tsv":"'1':342,368,407,415,444,742,760,823,882,903,1069 '10':654 '2':339,375,492,628,883,920,930,939,1110 '20':655 '3':384,408,416,529,546,743,761,824,911,1140 '4':392,612,629,1177 '5':321,402,677,1213 '6':411,736 '7':422,790,944 '8':430,442,841 'abandon':907,926 'abil':667,1133 'ability/friction':38,568 'action':277,455,505,511,558,626,686,749,804,848,1188 'activ':86,904 'actual':1207 'add':675,678,857 'addict':180,1030 'address':1128 'advic':170,193 'agre':1169 'alway':436,894 'analysi':998 'analyt':309 'analyz':220,978 'anchor':1079 'annoy':954 'answer':327 'anti':1055 'anti-pattern':1054 'anxious':126 'appli':5,73,1063,1181 'aren':328 'artifact':60 'ask':318 'assumpt':334,557 'avail':330 'avers':586,1081 'aversion/progress':50,588 'avoid':1057 'b':567,699 'back':1007 'barrier':551,593,633,1138,1195 'barriers/drivers':36,387 'baselin':285,374,503 'becom':168 'behavior':2,6,12,17,26,34,43,74,76,164,210,274,343,351,377,379,385,412,449,464,477,481,549,590,795,825,865,913,1003,1010,1047,1064,1124,1147,1223 'behavioral-product-design':1,1009 'bend':701,961 'bend-not-break':960 'benefici':1046 'benefit':712 'bet':744,781,807 'bias':102,582,1071,1078,1097 'bias-name-drop':1070 'bias/friction/uncertainty':389 'bound':487 'boundari':972,1022 'box':301 'brainstorm':627 'brand/tone':299 'break':703,963 'broad':345 'build':65,418,793 'build-readi':417,792 'c':569,708 'card':421 'case':817 'celebr':964 'celebration/pause':649 'chang':1125 'channel':298 'chat':357 'check':479,532,597,660,725,778,830,869,951 'check-in':950 'churn':987 'cite':1096 'clarifi':456 'clear':784 'coercion':179 'cognit':1077 'cohort':224,981,1000 'collect':243 'come':1006 'commit':47,646 'common':1059 'complaint':545 'complet':88,523,1232 'concret':1102 'confid':754 'confirm':1157 'confirm-sham':1156 'connect':1099 'constraint':294,373,576,748 'context':41,268,284,369,452,459,474 'continu':732,1155 'control':1053 'cooki':1184 'cookie-cutt':1183 'cost':182,1153 'cover':21,763,1136 'creat':175 'current':495,501 'cutter':1185 'd':571 'daili':949 'dark':176,1040,1141 'data':225,245,1209 'day':945 'decept':178,730 'decis':58,837 'decision-readi':57 'default':46,641,1115 'defin':150,446,465 'defining-product-vis':149 'deliver':348 'design':4,10,19,30,42,89,204,227,236,248,256,353,413,595,691,796,826,867,1005,1012,1068,1201,1217 'designing-survey':247 'desir':24 'devic':48,647 'diagnos':33,1192 'diagnosi':386,550,591,624,923,1075 'diagnost':1001 'differ':764 'disguis':1143 'domain':188 'domain/legal':196 'draft':846 'drive':121 'driver':552 'drop':292,515,1020,1073,1106 'drop-off':514 'e':574 'e.g':45,580,1229 'easi':714 'econom':77 'edg':816 'effect':1228 'effort':521,755 'emot':518 'engag':232,491,993,1238 'engin':831 'ethic':54,709,723,1024 'even':288 'event':426 'everi':1095,1159 'evid':303,556 'exampl':901,902,938,973,1023 'execut':29,880 'exist':255,302 'expect':922,956 'experi':15,31,423,798,921,935 'experiment/instrumentation':828 'explain':1175 'explicit':333,690 'failur':765,1060 'feasibl':840 'file':362 'final':844,864 'financi':192 'first':144,618,622 'flow':80,207 'flow/jtbd':502 'forc':731,1154 'forget':1129 'forward':135 'frame':645,710 'friction':530,534,561,666,767,1089,1211 'funnel':504 'funnel/retention':286 'gate':843,886 'generat':613 'generic':169,1178 'get':106,471 'goal':172,371 'grace':706 'ground':1204 'guardrail':55,383,467,821,970,1215 'habit':40,92,946,1179,1200 'habit/reinforcement':957 'habit/retention':688 'happen':293 'help':71,1171 'hesit':1109 'hidden':181,734,1152 'higher':1237 'hypothesi':611,785,812 'idea':395,615,656,664,751 'identifi':563 'impact':753 'implement':833 'implic':596 'improv':84 'in':952 'in-chat':355 'incl':959 'includ':437,895 'increas':1220,1230 'info':316 'input':264,450,500,553,623,683,745,800,845 'instrument':424,818,936 'intak':445 'intervent':13,44,228,393,404,420,614,630,652,684,720,727,810,928,931,1004,1114,1160,1218 'inventori':653,685,747 'journey':346,398,496,525,554,1094,1208 'know':1015 'label':336,610 'least':603,662 'len':212 'librari':637 'like':578,705 'list':1076 'loop':93,689,947,1180,1190 'loss':49,587,1080 'lower':1234 'major':835 'make':1026 'manipul':682 'map':14,394,396,493,526,577,929,1085 'mark':513 'markdown':359 'matter':499 'meaning':697 'measur':787,838 'mechan':400,579,594,600,617,658,813 'mechanism-first':616 'medic':191 'metric':166,222,287,382,428,788,971,997,1216 'minimum':265 'minor':194 'miss':315,836,1214 'missing-info':314 'mode':766,1061 'moment':399,497,519,555,650,695,927,965 'motiv':37,771,1132 'motivation/benefit':565 'move':133 'movement':789 'must':469,1098,1161,1202 'name':159,1072 'narrow':341 'need':82,141,202,218,234,251,976,996 'new':905 'next':434,861,899 'note':306,724 'notif':297,955,1112,1118,1127 'notification-as-intervent':1111 'nudg':1145 'observ':483 'off':516 'offer':338 'onboard':87,206,216,917 'one':461,604,663,669 'open':432,859,897 'opt':716,1051 'opt-out':715,1050 'order':367 'outcom':512 'output':347,473,524,589,651,718,772,822,863 'pack':354,847,868,871 'pass':1162 'pattern':177,636,1041,1056,1142 'paus':694 'penalti':735 'peopl':112,1032 'per':631 'percept':566,573 'period':707 'pick':739,759 'plan':16,32,425,799,829,937,958 'platform':295 'pm/design/eng':62 'point':531,535,562,1090,1212 'polici':704 'posit':146 'present':101,581 'primari':462,1123 'primary/guardrail':427 'priorit':154,403,737,773 'prioriti':632 'prioritizing-roadmap':153 'proceed':331 'procrastin':100 'produc':56,349 'product':3,9,18,151,267,352,866,876,1011,1067 'progress':698 'progress/loss':644 'prompts/forgetting':570 'proof':1083 'propos':599,919 'push':1117 'qualiti':842,885 'question':322,433,860,898 'quo':117,584 'rational':410,777 'readi':59,419,794 'real':258 'redesign':916 'redirect':974,989 'reduc':51,130,642,665,670 'references/checklists.md':850,851,889,890 'references/intake.md':324,325,453,454 'references/rubric.md':854,855,892,893 'references/templates.md':439,440,802,803 'references/workflow.md':639,640 'refram':1042 'refus':1025,1039 'regulated/high-stakes':190 'reinforc':680,693 'reinforcement/resilience':722 'reli':728 'remind':94,648,676,1120 'replay':311 'request':364,995 'requir':195,266,887 'research':240,305 'research/quote/data':607 'resili':679,700 'respons':988,1038 'retent':85,231,980,992 'retention-engag':230,991 'retention/engagement':221 'retention/habit':940 'review':197 'reward':1189 'risk':401,431,757,858,896 'roadmap':147,155 'rollout/rollback':429,819 'rough':289 'run':547,849 'safeguard':200 'satisfact':1235 'scienc':7,75,211,914,1065 'scope':20,340 'score':750,852 'segment':271,372,1199 'select':780 'session':310 'setup':908 'shame':1158 'shortlist':405,774,801 'show':390 'side':1227 'signal':606 'sketch':506 'skill' 'skill-behavioral-product-design' 'snapshot':370,475 'social':1082 'social/context':575 'source-liqiongyu' 'spammi':98 'spec':378,414,478,797,811,827,932,968 'specif':540,873,1088,1194 'sprint':884 'state':815 'statement':380 'status':116,583 'step':435,443,508,862,900,910,1103 'step/state':541 'stick':113 'stop':1035 'strategi':143,317 'streak':91,967 'studi':241 'success':165,381,466 'summari':527 'support':307,605,1241 'survey':238,249 'switching/adoption':122 'system':95 'tabl':592,625 'tag':659 'target':11,161,163,269,273,376,448,463,476,480,1222 'task':1231 'team':63 'techniqu':1148 'templat':438 'test':67,253,263,1165 'thing':111 'ticket':308,1242 'tie':537 'tighter':199 'time':300,486 'time-bound':485 'tool':1126 'top':406,528,741 'topic-agent-skills' 'topic-ai-agents' 'topic-automation' 'topic-claude' 'topic-codex' 'topic-prompt-engineering' 'topic-refoundai' 'topic-skillpack' 'toward':1043 'track':1225 'transpar':713,1048,1164 'trick':1150 'trigger':510,1187 'trigger-action-reward':1186 'trust':672 'trust/legal/brand':758 'turn':22 'type':1139 'ui':621,1028 'ui-first':620 'uncertain':125 'uncertainti':39,52,131,520,585,643,671,769,1135 'uncertainty/risk':572 'understand':983 'unintend':1226 'unknown':337 'updat':719,746 'upstream':142 'usability-test':261 'usabl':262 'use':35,70,139,148,213,229,246,260,634,888,912,1036,1146 'user':25,99,123,162,215,244,259,270,276,304,451,458,711,906,985,1019,1045,1093,1105,1151,1168,1198 'user-benefici':1044 'user-onboard':214 'ux/copy':814 'vagu':544 'vision':145,152 'vs':344,768,770 'wait':522 'want':279,942 'web/mobile':296 'without':96,208,681,834,953,1074,1084,1191,1224 'workflow':441 'wors':472 'would':1166 'write':791,808","prices":[{"id":"9e8525fd-c3d5-4f70-86ba-94dda37602aa","listingId":"47056ebf-1944-486f-b89e-b3d5874d4a8c","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"liqiongyu","category":"lenny_skills_plus","install_from":"skills.sh"},"createdAt":"2026-04-18T22:16:14.512Z"}],"sources":[{"listingId":"47056ebf-1944-486f-b89e-b3d5874d4a8c","source":"github","sourceId":"liqiongyu/lenny_skills_plus/behavioral-product-design","sourceUrl":"https://github.com/liqiongyu/lenny_skills_plus/tree/main/skills/behavioral-product-design","isPrimary":false,"firstSeenAt":"2026-04-18T22:16:14.512Z","lastSeenAt":"2026-04-22T06:56:18.784Z"}],"details":{"listingId":"47056ebf-1944-486f-b89e-b3d5874d4a8c","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"liqiongyu","slug":"behavioral-product-design","github":{"repo":"liqiongyu/lenny_skills_plus","stars":49,"topics":["agent-skills","ai-agents","automation","claude","codex","prompt-engineering","refoundai","skillpack"],"license":"apache-2.0","html_url":"https://github.com/liqiongyu/lenny_skills_plus","pushed_at":"2026-04-04T06:30:11Z","description":"86 agent-executable skill packs converted from RefoundAI’s Lenny skills (unofficial). Works with Codex + Claude Code.","skill_md_sha":"a0a62d6916e9fe675561bfa22e70d7f96869bc73","skill_md_path":"skills/behavioral-product-design/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/liqiongyu/lenny_skills_plus/tree/main/skills/behavioral-product-design"},"layout":"multi","source":"github","category":"lenny_skills_plus","frontmatter":{"name":"behavioral-product-design","description":"Apply behavioral science to product design: target behavior, intervention map, experiment plan."},"skills_sh_url":"https://skills.sh/liqiongyu/lenny_skills_plus/behavioral-product-design"},"updatedAt":"2026-04-22T06:56:18.784Z"}}