--- title: Customer research for ad creative — generations, micro-moments, emotions slug: customer-research-for-ads summary: Sarah Levinger's customer-reality research (the "customer cake") — generational context (information, money, life stage, trust), micro-moments when the problem surfaces, and emotional mapping (valence/intensity, private vs social, acute vs chronic) — with where to find each, plus review-mining shortcuts from Dara Denney and Emily Hickey. domain: marketing tags: [ad-creative, customer-research, personas, consumer-psychology, creative-strategy] related: [creative-strategy-fundamentals, brand-reality-capacity-and-alignment, teep-funnel-stage-creative, motivator-map-why-buy-why-believe, human-desires-ad-angles, ugc-engine-personas-and-creator-deals, voice-of-customer-copy-mining] sources: [yt:pV0tM1gSJiU, yt:t-Xf12o4jt4, yt:ILTXC1L8te8, yt:ZydGeio1DGQ] updated: 2026-10-24 --- # Customer research for ad creative ## TL;DR - Sarah Levinger's system has **three realities**: customer (how they experience the problem), brand (what the business needs now — [[brand-reality-capacity-and-alignment]]), presentation (which format the brain in that state accepts — [[teep-funnel-stage-creative]]).[^1] - Research for **insights, not fun facts**: "she wants to lose weight" is a fact; "getting dressed for lunch with a friend who just lost 40 lb and her jeans don't fit" is an insight that tells you creator, tone, brief.[^2] - Customer research has 3 layers, in order: **generational context → micro-moments → emotional mapping**.[^3] - Small team? Skip generational; do micro-moments, emotions and formats.[^4] - Fastest shortcut: upload all reviews as CSV and search for **failed attempts** before your product (feeds "investment" hooks).[^5] ## Details **Layer 1 — generational context (4 questions).** The brain maps new stimuli onto past experience; familiar formats get processed with less resistance (familiarity effect) — e.g. boomers raised on magazines respond to long-form article-style ads.[^6] 1. Childhood information environment (3 TV channels and authority vs infinite internet and "everyone has an agenda") → which formats feel safe.[^6] 2. Relationship with money (90s boom vs graduating into 2008 and student debt) → "self-care starts at home" lands as aspiration or insult.[^7] 3. Actual life right now (building/pivoting career, sandwiched between parents and teens, newly retired) → relevance on a specific Tuesday.[^8] 4. How they judge risk and whom they trust → authority-averse audiences get peer/friend angles; authority-trusting get doctors/experts.[^9] Sources: Pew Research; age-specific subreddits, Facebook groups, Quora, LinkedIn; BLS and census data (category price shifts reveal hidden objections); YouTube comments, podcast reviews, Amazon Q&A for tone. Absorb vocabulary — don't write angles yet.[^10][^11] **Layer 2 — micro-moments.** A good strategist finds pain points; a great one finds **when** they show up (time of year, week, day) in real-world, not marketing, situations. Hyper-specific situations give you a place to put the camera, a creator, a story — and suit Andromeda's appetite for specific content.[^12][^13] Where: Reddit/Facebook groups/forums for first-person "I was at the dentist…" stories (near-ready scripts); Amazon/Trustpilot reviews (compare 5/3/1-star for objections); quizzes, CS tickets, post-purchase surveys for **why now**; TikTok/IG comments where people tag friends ("this is so me").[^14] **Layer 3 — emotional mapping.** For each moment ask: which emotion is strongest? private or social? acute or years-long? A private, chronic problem wants "you don't have to suffer alone anymore", not hypey "join others like you".[^15] Place emotions on Russell's (1980) **valence × intensity** model to pick hooks and stories.[^16] **Cadence.** First Monday of each month: refresh generational, micro-moment and emotion research per avatar.[^17] **Other research moves.** - Emily Hickey's best headline source: **customer and non-customer interviews** (e.g. "I stopped wearing jeans at 40 because they're not comfortable"); also founder/design-decision interviews, ad and site comments, listicles ("finally, a turtleneck that doesn't choke your neck").[^18] - Dara: Ads Manager demographic breakdowns every test and month, Motion's "intended audience" AI tag, and plain stalking of customers on social.[^19] - AI: Sarah stopped asking LLMs to "do research" (they echo your bias); she has agents scrape real conversations by keyword, then asks AI what it sees.[^20] ## Disagreements / caveats - Sarah calls avatars "a crutch" — scale comes from a distinct brand point of view applicable to any avatar; she prefers emotional avatars.[^21] Tushy and Dara still name and rank personas (see [[ugc-engine-personas-and-creator-deals]]). ## See also - [[human-desires-ad-angles]] · [[motivator-map-why-buy-why-believe]] - [[people/sarah-levinger]] [^1]: yt:pV0tM1gSJiU @ 03:05 — Sarah Levinger [^2]: yt:pV0tM1gSJiU @ 09:12 — Sarah Levinger [^3]: yt:pV0tM1gSJiU @ 10:42 — Sarah Levinger [^4]: yt:pV0tM1gSJiU @ 1:21:18 — Sarah Levinger [^5]: yt:t-Xf12o4jt4 @ 00:00 — Dara Denney [^6]: yt:pV0tM1gSJiU @ 12:12 — Sarah Levinger [^7]: yt:pV0tM1gSJiU @ 13:43 — Sarah Levinger [^8]: yt:pV0tM1gSJiU @ 15:14 — Sarah Levinger [^9]: yt:pV0tM1gSJiU @ 16:46 — Sarah Levinger [^10]: yt:pV0tM1gSJiU @ 19:52 — Sarah Levinger [^11]: yt:pV0tM1gSJiU @ 21:27 — Sarah Levinger [^12]: yt:pV0tM1gSJiU @ 23:01 — Sarah Levinger [^13]: yt:pV0tM1gSJiU @ 24:35 — Sarah Levinger [^14]: yt:pV0tM1gSJiU @ 26:05 — Sarah Levinger [^15]: yt:pV0tM1gSJiU @ 27:38 — Sarah Levinger [^16]: yt:pV0tM1gSJiU @ 29:10 — Sarah Levinger [^17]: yt:pV0tM1gSJiU @ 1:13:31 — Sarah Levinger [^18]: yt:ILTXC1L8te8 @ 16:42 — Emily Hickey [^19]: yt:ZydGeio1DGQ @ 09:00 — Dara Denney [^20]: yt:pV0tM1gSJiU @ 1:33:57 — Sarah Levinger [^21]: yt:pV0tM1gSJiU @ 1:26:02 — Sarah Levinger