About关于我 Method方法 Projects项目 Contact联系
Insight · Product · Growth 洞察 · 产品 · 增长

Jackie Hu

胡佳琪

Consumer Insight & Product Strategy

消费者洞察与产品策略

I find out what customers will actually pay for, and turn it into decisions a product team can build. Automotive, luxury and FMCG.

我负责搞清楚客户到底愿意为什么付钱，再把它变成产品团队能落地的决定。汽车、奢侈品、快消。

View Projects 查看项目 LinkedIn
Since 2020
In the Field
从业至今
F500
Client Tier
世界500强
0
Languages
掌握语言
About关于我

Between what a brand intends and what a market does. 在品牌的意图，和市场的反应之间。

A brand arrives with an intention — a positioning, a standard, a thing it wants to mean to people. The market responds on its own terms, which are rarely the same terms. My work sits between the two: reading what customers are actually responding to, and turning that into something a product or design team can build.

品牌总是带着一个意图过来——一套定位、一个标准、一个它希望自己在人们心里成为的样子。而市场按自己的逻辑回应，这两套逻辑很少对得上。我的工作在这两者中间：读懂客户真正在回应什么，再把它变成产品或设计团队能落地的东西。

How I Work
方法与生态
Research Methods 研究方法
In-Store Observation门店行为观察 Intercept & In-Depth Interviews拦访与深度访谈 Segmentation & Personas消费者分群与画像 Competitive Experience Walkthroughs竞品体验走查 Journey Mapping用户旅程地图 Survey Design问卷设计 Funnel & Behavioural Analytics漏斗与行为数据分析 AI-Assisted SynthesisAI 辅助编码与综合
China Commerce Ecosystem 中国商业生态
WeChat RED Douyin Tmall JD.com
Commercial Toolkit 商业工具箱
Product Positioning产品定位 Conversion Funnel Design转化漏斗设计 Pre-Sales & Solution Design售前方案设计 CRM / SCRM Dashboards & KPI数据看板与 KPI UAT & Vendor ManagementUAT 与供应商管理 AR / AI Solutions
Languages
语言能力
中文
Native 母语
English
Advanced 高级
Français
Convers. 日常交流
Method方法

Where the evidence has to come from. 证据该从哪里来。

Different questions need different evidence. A positioning call and a usability problem are not answered the same way — most of the work is deciding which quadrant a brief actually sits in.

不同的问题需要不同的证据。定位判断和易用性问题，本来就不该用同一种方式回答——大部分功夫花在判断一个需求究竟落在哪个象限。

THE MARKET · category, competitors, size 市场侧 · 品类 · 竞争 · 规模 THE PERSON · behaviour, motivation, meaning 个体侧 · 行为 · 动机 · 意义 DEPTH 深度 SCALE 规模 Competitive experience walkthroughs Expert & stakeholder interviews Trend and desk research Market and category scanning Segment sizing Competitive benchmarking Surveys Funnel & behavioural analytics A/B testing In-context observation Intercept & in-depth interviews Segmentation from behaviour

Most of my work has started in the lower-left quadrant — depth on the person — because the briefs I get tend to arrive before anything exists to measure. The rest of the map is what makes that evidence usable: sizing tells you whether a finding is worth acting on, competitive work tells you whether the position is already taken, and analytics tell you afterwards whether the call was right. A finding without the other three quadrants is an anecdote. 我的工作大多从左下角开始——个体侧的深度——因为需求到我这里时，通常还没有可以度量的东西。地图上其余三块，是让这份证据变得可用的部分：规模告诉你一个发现值不值得动手，竞争告诉你这个位置是不是已经被占了，行为数据在事后告诉你当初判断得对不对。只有一个象限的发现，是轶事，不是结论。

Selected Work精选案例

Projects that moved the needle. 真实落地，数据说话。

THE DESIGN PROCESS · WHERE RESEARCH ENTERS 01 Design brief who is this car for · what must it say · what it must cost 02 Themes many sketch directions, narrowed by the studio to a few 03 Theme evaluation surviving themes shown to the target group, against competitor cars — this is where the shape gets decided 04 Clay model · Class A the chosen theme made physical, then surfaced 05 Design freeze → show car the position, built once, put on a stand MY SCOPE Setting the brief dealership observation → which group this car is for, and what that group is actually buying everything downstream — themes, clay, freeze — inherits this decision. Get it wrong and no clinic recovers it.
JAC · 江淮 Show Car
Design Research · Setting the Brief
01 / 04 · Automotive · Show Car
01 / 04 · 汽车 · 概念车

JAC Show Car — Deciding Who the Car Is For 江淮概念车 · 决定这台车为谁而造

Studios do not run short of ideas — they run short of grounds for choosing between them. My work was deciding who this car was for, and what that group actually buys. Everything downstream inherits that call.

设计室从不缺想法，缺的是在想法之间做选择的依据。我做的是定这台车为谁而造，以及那群人真正在买什么。后面所有阶段都继承这个判断。

EvidenceDealership observation and intercepts, then a survey to size the groups. Three emerged — but a show car is one object, so two had to be given up.门店观察与拦访，之后用问卷给各组估规模，得到三组。而概念车只有一台，必须放弃两组。
Into the briefThat group's expectations written as attributes a designer can work against — stance, surfacing, lighting, CMF — not a persona.那组人的期待，写成设计师能对着做的属性——姿态、形面、灯语、CMF，不是 persona。
Why it mattersA clinic can only pick the best of what it is given. If the brief points at the wrong group, nothing later recovers it.clinic 只能在给它的东西里挑最好的一个。brief 指错了人群，后面补不回来。
//
A show car does not go to market — it is the cheapest way to put a positioning in front of people before tooling money is committed. The honest test is not applause, but whether the group you aimed at recognises itself. 概念车不上市——它是投模具钱之前，把一个定位摆到人面前最便宜的方式。诚实的验收标准不是掌声，是你瞄准的那群人有没有认出自己。
Design Brief Research Contextual Observation Segmentation Positioning
Kering Pavilion at CIIE 2025
Kering · 开云集团
AR Gamification · CIIE 2025
02 / 04 · Luxury · Exhibition
02 / 04 · 奢侈品 · 国家级展会

Kering Pavilion at CIIE 2025 — AR Treasure Hunt 2025 进博会开云馆 · AR 寻宝互动体验

Ten Houses, one pavilion, six days. The problem was not the installation — it was what this pavilion was for once the doors closed. We designed the hardware as well as the software, so the position had to be settled before anything was built.

十个品牌、一个展馆、六天。要解决的不是装置本身，而是门关上之后，这个展馆到底为了什么。硬件和软件都是我们设计的，所以定位必须在动工之前就定下来。

Choosing the positionA spectacle that photographs well, a giveaway that gets taken home, or a route that ends in a contactable trace. Only the third survives the doors closing — so the hardware was designed around movement and capture, not around looking impressive.拍照好看的奇观、能被带走的赠品、或一条以"可触达的痕迹"收尾的动线。只有第三种在门关上之后还存在——所以硬件是围绕动线与捕获设计的，不是围绕"看起来震撼"。
EvidenceRequirement interviews across the brand teams; on-site observation of how visitors route and where they stall.跨品牌团队的需求访谈，加上现场观察观众实际怎么走、在哪里停滞。
Design consequenceTen Houses shared one visitor need — a reason to keep walking. The interface became the guide, and every stop had to pay the visitor back.十个品牌共享同一个观众需求：一个继续往前走的理由。界面因此成了导览员，每一站都必须给观众回报。
//
Conversion logic built into the experience: every AR interaction ends in a shareable, trackable WeChat touchpoint — expo foot traffic converted into owned social reach and private-domain assets, not just dwell time. 转化逻辑内嵌于体验：每次 AR 互动都以可分享、可追踪的微信触点收尾——把展会公域客流沉淀为品牌自有传播与私域资产，而非止步于"停留时长"。
0+
Houses in
One Journey
品牌空间
一线串联
1-Tap
Share一键
分享
To WeChat
直达微信
Expo →
Private公域 →
私域
Traffic Conversion
客流沉淀
Stakeholder Research Visitor Observation Positioning Hardware + Interface
WATCH FIT 5 · drag to rotate — live production build WATCH FIT 5 · 拖动旋转——线上真实交付版本
HUAWEI Mate 80 Pro
MATE 80 PRO · view official page ↗ MATE 80 PRO · 查看官网页面 ↗
HUAWEI FreeClip 2
FREECLIP 2 · view official page ↗ FREECLIP 2 · 查看官网页面 ↗
Huawei · 华为
3D Interactive Web
● Live 3D
03 / 04 · Consumer Tech
03 / 04 · 消费科技

Global 3D Interactive Product Experiences 华为全球官网 3D 互动产品页

Delivered web-based interactive 3D product experiences for Huawei's global consumer sites — Mate 80 Pro, FreeClip 2 and WATCH FIT 5 — embedded at the point of purchase on official commerce pages across international markets. Owned the requirements-to-launch pipeline: specs, engineering and vendor alignment through UAT, and multi-market rollout.

为华为全球消费者官网交付 3D 互动网页——覆盖 Mate 80 Pro、FreeClip 2、WATCH FIT 5 多条产品线，直接嵌入各国官网商城购买页的成交环节。端到端负责从需求到上线：规格文档、工程与供应商协同 UAT、多市场发布。

The questionInteractive 3D is expensive and slow to load. On which products does it earn that back, and where is a cheaper format better?3D 交互开发贵、加载慢。哪些产品上它赚得回来，哪些产品上更便宜的形式反而更好？
Deciding testDoes a physical detail carry the purchase argument? A watch is bought on how it sits on a wrist; a still image cannot settle that. Elsewhere rotation adds cost without changing the decision.判断依据：购买理由是不是由某个物理细节承载的。手表买的是"戴在手上什么样"，静态图定不了；其他位置上"能转"只增加成本、不改变决策。
Role of the interfaceHere the interface is the sales floor — no showroom, no salesperson. Interaction design is not decoration on the page; it is the part doing the selling.在这里界面就是销售现场——没有展厅，没有店员。交互设计不是页面上的装饰，它就是真正在卖货的那部分。
//
The 3D window in this card is the actual production deliverable, streamed live from the official deployment — not a mockup. 卡片中的 3D 窗口即为线上生产环境的真实交付物，直连官方部署——不是演示稿。
0
Product Lines
Delivered
产品线交付
Global
Rollout全球
多市场
Multi-market Sites
多国官网上线
In-Page
3D官网
内嵌
Commerce Embedded
购买页集成
Comparative Evaluation Interaction Design Cross-device Validation
M&M's Disneytown Flagship Opening, Shanghai
Mars · M&M's 玛氏
WebAR · O2O Retail
04 / 04 · FMCG Retail
04 / 04 · 快消零售

M&M's Disneytown Flagship — WebAR O2O Funnel M&M's 迪士尼小镇旗舰店 · WebAR O2O 转化漏斗

For the launch of the world's first M&M's AR interactive experience store — the opening that made Shanghai the brand's only "dual-store city" — delivered an interactive WebAR game and in-store installation wired into an O2O acquisition funnel: play → scan → member registration → CRM asset. Data dashboards tracked every step against commercial KPIs.

全球首家 M&M's 品牌 AR 互动体验门店在迪士尼小镇启幕，上海由此成为品牌全球唯一"双店城市"。交付互动 WebAR 游戏与门店装置，并接入 O2O 获客漏斗：游戏互动 → 扫码 → 会员注册 → 沉淀为品牌 CRM 资产；数据看板全程对照商业 KPI 追踪每一步转化。

The questionFootfall in a tourist location is rented; a CRM record is owned. How much of one becomes the other?旅游地段的客流是租来的，CRM 记录才是自己的。前者能转成多少后者？
Where to askEntrance turns it into a toll gate; exit arrives after the visitor has mentally left. In-store observation located the third option.放入口变成收费闸，放出口人已心理离场。是店内观察定位出了第三个选项。
FindingDwell time and conversion do not move together. Willingness peaks at the emotional high point, not at the end — so the trigger and its wording moved there.停留时长和转化不是一起走的。意愿峰值在情绪高点、不在结束时——触发点与话术据此前移。
//
Retail entertainment engineered as an acquisition channel — a 2-minute average session converted foot traffic into first-party member data the brand owns and can re-market to. 把零售娱乐设计成获客渠道——平均 2 分钟的互动停留，将线下客流转化为品牌自有的一方会员数据，可持续复购触达。
0min分钟
Avg. Session
Duration
平均互动
停留时长
O2O →
CRMO2O →
会员
Acquisition Funnel
获客漏斗
Only
Dual-City唯一
双店城市
Global First
全球首例
In-store Observation Behavioural Metrics Conversion Research