AI与用户研究实战
- Chris Zhang
- Nov 7, 2024
- 24 min read
Updated: Jan 17
Why User Research Matters
User research is like giving your product a pair of "X-ray glasses" that let you see straight into the user's soul. It not only helps you figure out what users really want, but also reveals what’s going on inside their heads—those little thoughts and potential pain points you might never have noticed. Now, think about it—there are so many electronic products out there, and if you only rely on wild creativity and tech wizardry, how could you possibly outshine companies that truly understand users' inner needs? A successful product isn’t just “pretty,” it’s about knowing how to make users feel great. Effective user research is like “knowing in advance” what users want, helping businesses create smart product strategies and beat competitors to the market. And if you do it right, users will even be like, “Wow, this product is like it was made just for me!”
用户研究就像是给产品穿上一双“透视眼”，让你看透用户的内心世界。它不仅帮助我们搞清楚用户真正想要什么，还能揭示他们心里到底在想些什么：那些你可能从来没注意到的小心思和潜在的痛点。现在，想想看，市场上那么多电子产品，如果你只靠天马行空的创意和技术，怎么可能比得过那些洞察用户内心需求的企业？成功的产品不光得好看、好用，还得懂得怎么让用户“会心一笑”。有效的用户研究就像是在提前知道用户想要什么，它能帮助企业提前制定聪明的产品策略，抢在竞争对手前面占领市场。
Think like a user, act with AI. 像用户一样思考，借助AI行动。
Course Overview
This course is like giving you a cheat code! We’re going to take AI and user research to the next level, helping you acquire a practical skill set that’s both comprehensive and grounded. Not only will we teach you how to use AI tools to breeze through user research, but we’ll also help you understand user psychology, read between the lines, and provide strategic insights to guide product development—ultimately, helping you make powerful moves!
For example, ChatGPT can become your little assistant and think tank in user research. Once you master the skill of “Prompt Engineering,” you can get high-quality insights from it. For instance, when analyzing user interview data, ChatGPT can quickly summarize key trends in feedback and extract user needs through customized prompts. What’s more, it can simulate user personas, rapidly validate interaction designs, or even generate more precise surveys and interview scripts. In short, AI in user research doesn’t just save time and effort, it makes your research results richer and more valuable!
Besides prompts, AI agents can help automate data categorization, identify trends, and even predict latent user needs and behaviors. You can have it quickly scan user comments or social media posts, perform sentiment analysis, and extract key focus areas to create more personalized product innovation strategies. Whether it’s a quick user test or a long-term research project, AI agents are your go-to for boosting efficiency and insights!
However, don’t get me wrong—user research theory still serves as the foundation for ensuring research quality, while AI tools act as a productivity amplifier. By combining core user research theories, we can clarify research goals, design scientifically rigorous methods, and ensure the reliability of results. The introduction of AI tools makes tasks like data processing, sentiment analysis, and user behavior prediction quicker and more accurate. So, only with the dual support of theory and tools can user research maintain depth and improve efficiency, avoiding the pitfall of relying solely on AI.
这门课就是给你开外挂！我们要把AI和用户研究这两大领域玩到极致，给你一套既系统又接地气的实战技能包。不仅教你用AI工具轻松搞定用户研究，还让你洞察用户心理，拿捏他们的小心思，助力产品开发，打出漂亮的一手牌！
比如，ChatGPT能成为用户研究的小助手与智囊团，只要掌握了“提示词工程”这项技能，就能让它为你提供高质量的洞察。比如，在用户访谈数据分析时，ChatGPT可以通过定制化的提示词，快速总结反馈中的关键趋势，提炼用户需求。不仅如此，它还能模拟用户角色，进行交互设计的快速验证，或者生成更加精准的调查问卷和访谈脚本。总之，人工智能在用户研究中，不只是省时省力，还让你的研究结果更有料！
除提示词外，AI智能体能帮你自动化数据分类、识别趋势，甚至简单预测用户的隐性需求和潜在行为。你可以让它快速扫描用户评论或社交媒体帖子，进行情感分析、提炼关注点，从而制定更加个性化的产品创新策略。不管是短平快的用户测试，还是长周期的研究项目，AI智能体都是你提升效率和洞察力的好帮手！
不过，请不要误会，用户研究的理论框架仍是确保研究质量的基石，而AI工具则是放大生产力的催化剂。通过结合用户研究的核心理论框架，我们能够明确研究方向，制定科学的研究方法，并确保结果的可靠性。AI工具的引入，则让数据处理、情感分析、用户行为预测等复杂任务变得更快捷、更精准。所以，只有在理论和工具的双重支持下，用户研究才能既保持深度，又实现效率的提升，避免陷入单纯依赖AI的误区。
Our Clients
Feedback
As an international e-commerce business, we make a large number of decisions every day, such as which products to list, how to set pricing strategies, how to allocate advertising budgets, and so on. Many times, these decisions are based on experience and a sense of market trends. While there is some data support, most of the time, the adjustments are made based on intuition. However, the case studies in the course made me realize that AI can not only analyze historical data but also predict future trends and even uncover details we often overlook. Especially in market demand forecasting, AI can provide highly accurate predictions based on a vast amount of consumer data and trend analysis. This was truly eye-opening for me. In the past, I would make decisions based on superficial phenomena, which often led to "following the crowd" without digging into the underlying details. Now, I'm thinking that if we could incorporate these technologies into our development process to predict which features would be more popular or identify blue ocean markets, it would greatly reduce my trial-and-error costs. After all, with such fierce competition in the market, blindly producing and selling products without precise targeting will only cause us to miss opportunities in the intense competition. I can say this course exceeded my expectations and has started making me reflect. Initially, when I heard that AI applications could bring change, I had some doubts. First, I doubted the effectiveness of AI; second, I doubted my own ability because I had no idea what deep learning or machine learning meant. After completing the course, my feeling was that I had opened the door to a new world. Although the course content seemed quite advanced, the way it was taught was very practical, especially since the course was combined with real-world case studies, directly highlighting the practical applications of AI in e-commerce, showing me the direct value that technology brings.
作为跨境电商，我们每天都在做大量的决策，包括哪些产品上架、定价策略怎么制定、广告怎么投放等等，很多时候这些决策是基于经验和市场趋势的感觉做出来的。虽然有一些数据支持，但基本上还是凭直觉去调整。 然而，课程中的案例让我意识到，AI不仅能分析历史数据，还能预测未来的趋势，甚至能找到一些我们常常忽略的细节。尤其是市场需求预测这一块，AI能基于大量的消费数据和趋势分析，给出非常精确的预测结果。这对我们来说，真的是眼前一亮。以前我总是根据一些表面现象来做决策，结果往往是“跟风”，而没有深挖到底层。 我现在在想，如果能把这些技术引入到我们开发过程中，通过数据来预测哪些功能更受欢迎，蓝海市场在哪里，这将极大地减少我的试错成本。毕竟，市场上竞争这么激烈，盲目生产和销售没有精准定位的产品，只会让我们在激烈的竞争中失去机会。 可以说，这门课程有点超出预期，也开始让我反思。最开始，听到AI的应用能够带来改变时，我心里其实有点怀疑，一是怀疑AI的效果，二是怀疑我的能力。因为我根本不懂什么叫深度学习或者机器学习这些复杂的术语。参加完课程后，我的感觉是有点打开了新世界的大门，虽然课程内容看起来挺高大上的，但讲解方式很实用，尤其是课程结合了真实案例，直接指出AI在电商中的实际应用，让我看到了技术带来的巨大价值。
To be honest, after the course, I had a sudden moment of clarity and was deeply inspired. I have also clarified the direction for future product development. I will definitely start with the 'Minimum Viable Product' and spare no effort in improving 'Product-Market Fit.' I think the first thing I will do when I return to the company tomorrow is to review and summarize the knowledge learned in the course and use AI to analyze user pain points and product direction. I rarely used AI before, but now I realize that only through continuous practice and familiarization can one become proficient. Actually, I had heard of Chris’s training a long time ago and had several friends recommend it, so I started this learning journey with high expectations. The entire course exceeded my expectations. The two days of lessons started on time, with only a 10-minute break between the morning and afternoon sessions, but I didn't feel tired at all. I have a lumbar disc herniation and cannot sit for long periods. I have taken other courses before and found it hard to sit still, but this time, I forgot about the pain and was fully immersed in the learning process. This course has been very rewarding, and Chris’s professionalism and humor left a deep impression on me.
说实话，学习后，茅塞顿开，深受启发。我也明确了以后产品开发的方向。一定先从“最小化可行产品”出发，不遗余力地提高“产品市场契合度”。我想，明天我回到公司第一件事就是先复盘和总结课程中学习到的知识，用AI梳理用户痛点和产品方向。我之前很少使用AI，现在发现，只有不断练习，不断熟悉，才能熟能生巧。其实，Chris老师的课程早有耳闻，也有多个朋友推荐，所以这次是带着很高的期待开始的学习之旅。整个课程下来，远远超出了我的预期。两天的课程准时开课，上午和下午，中间都只有10分钟的休息时间，却一点没感觉到累。我本身自己患有腰椎间盘突出，不能久坐。之前也上过其他课程，很难坐得住。这次却让我忘却了疼痛，全身心投入了课程学习中。此次学习也让我收获满满，Chris老师的专业和幽默也给我留下了深刻的印象。
After attending Chris's training, the sentence that left the deepest impression on me was: "Instead of creating something that everyone understands a little, it's better to create something that a few people can truly understand and enjoy, and dominate that market." Therefore, I highly recommend it! I can honestly say that Chris provided us with a foundational understanding of user research and a complete methodology. I now truly appreciate why Apple has become the industry's benchmark. It was truly enlightening. I deeply admire and respect Chris for his thoughtful approach. After two days of training, I realized that this might be the underlying logic behind Apple's product development. It has completely reshaped my thinking and left me in awe. I plan to carefully reflect on and digest this knowledge, and make corresponding attempts and changes. Oh, by the way, I really liked how Chris used the "King of Hearts" analogy to explain the sunk cost fallacy—how product managers often fail to create great products yet remain unaware of their own self-deceptive trap.
上完Chris老师的课，让我印象最深的一句话就是：与其做一个每个人都理解一点的东西，还不如做一些少数人能深刻领会、感到喜悦的东西去垄断那个市场。所以，我满分推荐！可以说Chris老师给我们带了用户研究的底层认知和完整的方法论。我现在真的能体会到，平苹果公司为什么能成为行业的天花板。真的是大受启发。非常欣赏和敬佩Chris老师的良苦用心。上完两天的课，我想，这大概就是苹果公司做产品的底层逻辑吧，刷新了我的认知和思考，深感震撼，我想我会好好思考和消化这些知识，并做出相应的尝试和改变。哦，对了，我真的很喜欢Chris老师课上用“红桃K”比喻沉没成本偏差 - 产品经理做不出好产品却又毫不自知的自证陷阱。
Who Should Attend
User experience designers, product managers, data analysts, and other professionals looking to enhance their user research skills.
Chief product officers and CEOs who aim to leverage user research to create product differentiation and enhance competitive advantage.
Professionals interested in artificial intelligence technology and eager to apply AI to solve problems in their work.
Team members from startups who want to conduct efficient user research with limited resources.
用户体验设计师、产品经理、数据分析师等希望提升用户研究能力的职场精英。
企业产品一号位和CEO，希望通过用户研究打造产品差异化、提升产品竞争力。
对人工智能技术感兴趣，并希望在实际工作中应用AI解决问题的专业人士。
初创公司团队成员，希望在资源有限的情况下高效开展用户研究。
Course Structure
Course Duration: Two days, from 9:00 AM to 6:00 PM each day, fully offline.
Teaching Method: In-person classes with real-time interaction and hands-on practice.
课程时长：两天，每天早九点到晚六点，完全线下授课。
授课形式：线下面对面教学，实时互动与实践操作。
关键课题
The Origins and Development of User Research
In the first module, Chris will delve into the development of user research and its significance in product innovation design. We will begin by reviewing the background of user research and its relationship with market research. Market research has evolved from traditional surveys to big data analysis, playing a crucial role in optimizing product strategies. The emergence of User Experience (UX) marks a shift from mere functional design to a user-centered, comprehensive experience design. User research considers user behavior, psychology, and socio-cultural contexts, ensuring seamless alignment between products and user needs. Additionally, Chris will share insights from psychology on the importance of the "unconscious" in marketing and product design, and discuss the value of "pioneer users," including extreme and expert users, in product innovation. By studying these topics, you will gain a comprehensive understanding of the origins and evolution of user research, laying a solid foundation for subsequent modules.
模块一：用户研究的起源与发展
在第一个模块，Chris老师会和大家深入探讨用户研究的发展历程及这门学科在产品创新设计中的重要性。 首先，回顾用户研究的背景，了解其与市场研究的关系。市场研究是从传统的市场调研到大数据分析，在产品策略优化中发挥了重要作用。而用户体验（UX）的出现，标志着产品从单纯的功能设计走向以用户为中心的全流程体验设计。用户研究综合考量用户行为、心理和社会文化背景，使产品与用户需求无缝衔接。此外，Chris老师还会从心理学角度与大家分享“无意识”在市场营销和产品设计中的重要性，分享“先锋用户”，即极端用户和专家用户对于产品创新的价值。 通过对这些内容的学习，相信您会对用户研究的起源和发展有了更全面的理解，为后续模块的学习奠定基础。
User Research is About Identifying Needs
The Essential Differences Between Qualitative and Quantitative Research：
Qualitative Research: Utilizes small-scale, carefully selected samples to explore user motivations, behaviors, and needs, focusing on understanding "why" and "how."
Quantitative Research: Employs large-scale samples to collect numerical data, validating hypotheses and concentrating on understanding "what."
Four Different Research Forms:
Complex Research: Suitable for innovative projects with unclear needs and high trial-and-error costs.
Simple Research: Suitable for innovative projects with unclear needs and low trial-and-error costs.
Design Evaluation: Suitable for innovative projects with clear needs and high trial-and-error costs.
Launch Measurement: Suitable for innovative projects with clear needs and low trial-and-error costs.
Five Levels of User Needs:
Explicit Needs: Clearly expressed needs by users.
Critical Needs: Core features that support user satisfaction.
Implicit Needs: Unspoken, hard-to-articulate yet real potential expectations.
Desire Needs: Users' imaginations and visions.
Blank Needs: Potential innovation points yet to be discovered.
模块二：用户研究的形式与意义
定性研究与定量研究的本质区别：
定性研究：通过小规模、精心挑选的样本，探索用户的动机、行为和需求，侧重于理解“为什么”和“怎么做”。
定量研究：通过大规模样本，收集数值数据，验证假设，侧重于了解“是什么”。
四种不同的研究形式：
复杂研究：适用于需求不清晰、试错成本高的创新项目。
简单研究：适用于需求不清晰、试错成本低的创新项目。
评估设计：适用于需求很清晰、试错成本高的创新项目。
上架测量：适用于需求很清晰、试错成本低的创新项目。
用户研究就是定义需求：
显性需求：用户明确表达的需求。
关键需求：支撑用户满意度的核心功能点。
隐性需求：用户未明说、很难一句话说清楚但又真实存在的潜在期望。
愿望需求：用户的想象与愿景。
空白需求：尚未被发现的潜在创新点。
LLMs (Large Language Models)
What Exactly Are Large Language Models Represented by ChatGPT?
Neural Network Principles: The capability of deep learning plays a crucial role in the performance of Large Language Models (LLMs).
Training Data Composition: The importance of text data sources, annotation methods, and quality control in training LLMs.
Attention Mechanism (Transformer Architecture): Efficient processing of long-sequence texts and understanding contextual relationships.
How LLMs Empower Product Teams:
Prompt Engineering: Enhancing the accuracy and practicality of model-generated content through meticulously designed prompts.
Knowledge Base Integration: Incorporating proprietary enterprise knowledge bases into models to improve the depth and professionalism of responses to domain-specific questions.
Intelligent Agents: Customizing intelligent agents to handle specific tasks, such as automatically generating questionnaires and analyzing interview texts.
Six-Step Method for Prompt Editing:
Role: Assigning a specific role to the model to enhance the professionalism of its responses.
Context: Providing a usage scenario to make responses more applicable to practical applications.
Task: Clearly defining the expected output (e.g., designing questionnaires, generating research hypotheses).
Format: Specifying the response format for subsequent information integration and use.
Examples: Providing reference examples to reduce response biases.
Tone: Setting an appropriate tone to improve the readability and friendliness of responses.
模块三：大语言模型
什么是大语言模型（ChatGPT）？
层和神经元： 深度学习能力对大语言模型（LLMs）的性能起到关键作用。
预训练数据： 文本数据来源、标注方式与质量控制的重要性。
注意力机制： 高效处理长序列文本，理解上下文关系。
人工智能如何赋能产品团队：
提示词： 提高模型生成内容的准确性与实用性。
知识库： 提高对领域问题回答的深度与专业度。
智能体： 处理特定、重复性、相对复杂任务，如自动生成问卷、分析访谈文本。
提示词撰写六步法：
角色： 为模型设定专角色，提升回答专业度。
场景： 提供使用场景，使回答更贴合实际应用。
任务： 明确期望模型完成的工作，如设计访谈问卷、生成研究假设等。
格式： 规定回答格式，以便后续信息整合与使用。
举例： 提供参考示例，减少模型回答偏差。
语气： 设定合适的语气，提高回答的可读性和亲和度。
Defining Target Users
Achieving PMF:
Product-Market Fit (PMF) : PMF refers to the degree to which a product satisfies a strong market demand. Achieving PMF means your product effectively addresses the needs of a specific market segment, leading to increased user adoption and engagement.
Minimum Viable Product (MVP) : An MVP is the simplest version of a product that allows a team to collect the maximum amount of validated learning about customers with the least effort. It includes only the core features necessary to test the product's value proposition and gather user feedback.
Minimum Viable Segment (MVS): MVS involves focusing on a specific market segment of potential customers with similar needs that your product can address. By identifying and targeting this segment, you can tailor your MVP to meet their unique requirements, increasing the likelihood of achieving PMF.
Three Strategies of Achieving PMF:
Hair on Fire: This archetype addresses problems that are urgent and critical for customers. The demand is immediate and evident, often leading to a crowded market with numerous competitors. To succeed in this space, a product must offer a differentiated and superior solution to stand out.
Hard Fact: This category involves solving problems that customers have accepted as a part of life, often due to inertia or lack of better alternatives. The challenge here is to convince customers that change is possible and beneficial. Innovative solutions can disrupt the status quo, leading to significant market opportunities.
Future Vision: This archetype focuses on visionary innovations that create entirely new markets or paradigms. Products in this category often face skepticism and require educating the market about the new possibilities they offer. While the path is challenging, the potential rewards are substantial.
Criteria for Determining Target Users (Made 4U):
Unworkable: Does the problem significantly impact the user?
Urgent: Is there an immediate need to solve the problem?
Unavoidable: Are users aware of the problem's existence?
Underserved: Can the product significantly enhance the user's experience and satisfaction?
Scientifically Recruiting Users:
Recruitment Screener: A recruitment screener is a tool—such as a document, survey, or call script—used to qualify participants for qualitative market research activities like focus groups, in-depth interviews, or in-home usage tests. It ensures that selected participants meet predefined criteria, aligning with the study's objectives and target demographics.
Quota Sampling: Quota sampling is a non-probability sampling method where researchers segment a population into distinct subgroups based on specific characteristics (e.g., age, gender, income). They then select participants non-randomly from each subgroup to meet predetermined quotas, ensuring that the sample reflects the population's composition concerning these characteristics.
Stratified Sampling: Stratified sampling is a probability sampling technique that involves dividing a population into mutually exclusive subgroups, known as strata, based on shared characteristics. Researchers then perform random sampling within each stratum to ensure that every individual has an equal chance of selection. This method aims to achieve a more precise and representative sample compared to simple random sampling.
模块四：定义目标用户
达成产品市场契合度（PMF）的基本原则：
产品市场契合度（PMF）：指产品满足强烈市场需求的程度，即您的产品有效地解决了特定市场细分的需求，从而提高了用户的采用率和参与度。
最小化可行产品（MVP）：以最少的资源和时间，推出具备核心功能的版本，以验证市场需求和获取用户反馈。
最小可行细分市场（MVS）： 创新的产品应专注于垄断具有相似（相同）需求的潜在用户特定细分市场，满足他们的独特要求，从而增加达成PMF的可能性。
达成产品市场契合度（PMF）的三条道路：
燃眉之急：解决用户迫切且明显的问题，需求紧急且显而易见，市场竞争激烈。要在此类市场中脱颖而出，产品需提供独特且卓越的解决方案。
木已成舟：解决用户已接受为生活常态的问题，用户对现状习以为常，缺乏改变的紧迫感。挑战在于说服用户接受新的解决方案，打破惯性思维。
未来愿景：通过前瞻性创新创造全新的市场或范式，产品可能面临用户的怀疑，需要教育市场，展示新可能性。尽管路径充满挑战，但潜在回报巨大。
确定目标用户的标准（Made 4U）：
痛不痛： 问题是否显著影响用户？
急不急： 问题是否需要立即解决？
看不见： 用户是否意识到问题的存在？
好不好： 产品能否显著提升用户的体验和满意度？
科学地招募用户：
招募标准：通常以调查问卷的形式出现，用于确定受访用户是否符合要求，与研究目标和产品目标人群一致。
配额抽样：一种非概率抽样方法，研究人员根据特定特征（需求、痛点）将总体划分为不同的子群体，然后在每个子群体中非随机地选择样本，以满足预定的配额要求，从而确保样本在这些特征上的组成与总体相符。
分层抽样：分层抽样是一种概率抽样技术，首先将总体根据共享特征划分为互不重叠的子群体（称为层），然后在每个层内随机抽取样本，以确保每个个体都有平等的被选中机会。
Hypothesis
Hypothetical Thinking: Hypothetical thinking is a cornerstone of product innovation, enabling teams to explore possibilities beyond current limitations and anticipate future needs. By posing "what if" questions, innovators can challenge existing assumptions, envision new product categories, and foresee emerging demands.
ChatGPT Prompt：
Defining Target Audience
Identifying Target Audience Pain Points
Defining Product Functional Requirements
Defining Product Emotional Requirements
Planning Product Ecosystem
Defining Product Interaction Journey
模块五：寻找产品假设
假设性思维：假设性思维是产品创新的基石，使团队能够超越当前的局限，预见未来的需求。通过提出“如果……会怎样？”的问题，创新者可以挑战现有假设，构想新的产品类别，预见新兴需求。
ChatGPT 提示词：
如何使用ChatGPT推理产品的目标用户？
如何使用ChatGPT识别目标用户的痛点？
如何使用ChatGPT定义产品的功能需求？
如何使用ChatGPT定义产品的情感需求？
如何使用ChatGPT帮助规划产品的生态体系？
如何使用ChatGPT推理产品的交互旅程？
Data Pattern
Triangulation: Triangulation is a research methodology that enhances the validity and reliability of findings by integrating multiple data sources, methods, and perspectives.
Survey Research: Collecting extensive user feedback through questionnaires or interviews to obtain both quantitative and qualitative data.
Descriptive Research: Providing detailed descriptions of phenomena, analyzing their characteristics and patterns, thereby laying the groundwork for subsequent studies.
Experimental Research: Controlling variables to observe experimental outcomes, testing hypotheses, and acquiring data on causal relationships.
How does ChatGPT assist in formulating interview questions?
How does ChatGPT assist in designing user research questionnaire?
What are the techniques for user interviews?
Crafting Role-Based User Personas:
Activities: Specific actions and tasks users perform to achieve their goals.
Environment: The context and setting in which these activities occur.
Interactions: Engagements between users and other individuals or systems during their activities.
Objects: Tools, devices, or items users interact with in their environment.
Users: Demographic and psychographic information about the users.
Formulating User Stories: User stories are concise, informal descriptions of software features from the end user's perspective. They outline the user type, their needs, and the reasons behind those needs, serving as placeholders for discussions about work to be done. The structure of a user story is: "As a [user type], I want to [perform an action] so that I can [achieve a goal]."
Formulating 'How Might We' (HMW) Questions: 'How Might We' is a common questioning technique in design thinking used to redefine challenges as opportunities for innovation. By starting questions with 'How might we', teams are encouraged to explore multiple potential solutions, fostering creativity and collaboration. The structure of an HMW question is: "How might we [action] so that we can [desired outcome]?"
模块六：收集数据、识别脉络
什么是三角测量法：通过结合不同的数据和研究方法提高研究结果的可靠性。
调查研究：通过问卷或访谈等方式，收集大量用户的反馈和意见。
描述研究：对现象进行详细描述，分析其特征和规律，为后续研究提供基础。
实验研究：通过控制变量，观察实验结果，验证假设，获取因果关系的数据。
如何使用ChatGPT帮助制定访谈问题列表？
如何使用ChatGPT帮助设计用户调研问卷？
用户访谈的技巧是什么？
角色型用户画像的撰写方法：
活动：用户为实现目标所进行的具体行为和动作。
环境：这些活动发生的整体环境和背景。
交互：用户为达成目标所进行的完整交互，包括人与人、人与产品之间的互动。
物体：产品使用环境中的所有细节，以及与用户、活动和交互的关系。
用户：受访用户的背景信息。
提炼用户故事：用户故事是从最终用户的角度对软件功能的简洁、非正式描述。它概述了用户类型、他们的需求以及需求的原因，充当了关于待完成工作的讨论的占位符。用户故事的结构：作为“用户类型”，我希望“执行某个操作”，以便“实现某个目标”。
提炼“How Might We”：“How Might We”意为“我们可以怎样”。 这是一种在设计思维中常用的提问方式，用于将挑战重新定义为创新的机会。 以“如何才能”开头的问题可以鼓励团队探索多种潜在解决方案，促进创造力和协作，有助于拓宽思维，避免陷入固定的思维模式，激发创新。
POC
Kano Model：The Kano Model, developed by Professor Noriaki Kano in the 1980s, is a framework for understanding customer satisfaction in product development. It categorizes product features into five distinct types based on their impact on customer satisfaction:
Must-Be Quality (Basic Needs): These are fundamental features that customers expect. Their presence doesn't significantly increase satisfaction, but their absence leads to dissatisfaction. For example, a smartphone's ability to make calls is a must-have feature.
One-Dimensional Quality (Performance Needs): Features that customers explicitly desire. The better these features perform, the higher the customer satisfaction. For instance, a car's fuel efficiency directly correlates with customer satisfaction. WIKIPEDIA
Attractive Quality (Excitement Needs): Features that delight customers when present but don't cause dissatisfaction when absent. These are unexpected perks that can significantly boost satisfaction. An example is a smartphone with a built-in projector.
Indifferent Quality: Features that neither enhance nor detract from customer satisfaction. Their presence or absence doesn't significantly impact the customer experience. For example, the color of a product's internal components might be indifferent to most customers.
Reverse Quality: Features that cause dissatisfaction when present and satisfaction when absent. This often occurs when a feature doesn't align with customer preferences. For example, a customer might prefer a simple, user-friendly interface over a feature-rich but complex one.
Storyboard: A storyboard is a sequence of illustrations or images that depict a user's journey through a product or service. In the context of user research, storyboards help visualize how users will interact with a PoC, highlighting potential pain points and areas for improvement. This approach fosters empathy and provides a clear understanding of the user experience.
Enhanced Communication: Storyboards provide a clear and engaging way to present ideas, making it easier for stakeholders to understand and discuss the concept.
Early Validation: By visualizing user interactions, storyboards help identify potential issues early in the development process, allowing for timely adjustments.
User-Centric Design: Focusing on user journeys ensures that the PoC is designed with the end-user in mind, leading to a more intuitive and satisfying experience.
Analyzing Quantitative Data with ChatGPT:
T-test: Used to compare the means of two groups to determine if there is a statistically significant difference between them. There are two types: independent samples T-test (for comparing two independent groups) and paired samples T-test (for comparing two related groups).
Analysis of Variance (ANOVA): Used to compare the means of three or more groups to determine if there is a statistically significant difference among them. One-way ANOVA is used for comparing one factor with multiple levels, while two-way ANOVA is used for comparing two factors simultaneously.
Chi-square Test: Used to analyze the association between categorical variables by comparing observed frequencies with expected frequencies. It is commonly used to assess whether there is a significant association between two categorical variables.
Correlation Analysis: Used to measure the strength and direction of the relationship between two or more variables. Pearson's correlation coefficient is used for continuous variables, while Spearman's rank correlation coefficient is used for ordinal variables or non-normally distributed continuous variables.
Regression Analysis: Used to examine the relationship between one dependent variable and one or more independent variables. Linear regression is used for continuous dependent variables, while logistic regression is used for binary dependent variables.
Principal Component Analysis (PCA): Used for dimensionality reduction by transforming a large set of variables into a smaller one that still contains most of the information. It is commonly used to identify the most important factors affecting user behavior.
Cluster Analysis: Used to group a set of objects into subsets or clusters, so that objects within the same cluster are more similar to each other than to those in other clusters. It is commonly used to identify different user segments.
Homogeneity of Variance Test: Used to test if different groups have the same variance. This is an important assumption for conducting T-tests and ANOVA.
模块七：概念验证
Kano模型：由东京理工大学教授狩野纪昭（Noriaki Kano）于1980年代提出，是用于理解产品开发中用户满意度的框架。该模型将产品特性分为五类：
基本型需求（Must-Be Quality）： 其存在不会显著提高满意度，但缺失会导致不满，例如智能手机的通话功能。
期望型需求（One-Dimensional Quality）：表现越好，用户满意度越高的需求，例如汽车的燃油效率直接影响车主满意度。
兴奋型需求（Attractive Quality）： 存在时能让用户感到愉悦，但缺失时也不会引起不满，例如内置投影仪的智能手机。
无差异需求（Indifferent Quality）： 其存在与否不会显著影响用户体验的需求，例如产品内部组件的颜色对大多数用户都无关紧要。
反向需求（Reverse Quality）： 存在时导致不满，缺失时反而则会令用户满意的需求，例如用户喜欢简单的界面，而非功能丰富的界面。
故事板：故事板描绘了用户在使用产品时的全部旅程。在用户研究中，故事板有助于可视化用户与产品概念（PoC）间的互动，突出潜在的痛点和改进领域。它的作用是：
增强沟通： 使用户充分理解创新产品的概念。
早期验证： 通过可视化用户互动，故事板有助于识别潜在问题。
以用户为中心的设计： 确保PoC的设计以最终用户为中心，专注于用户体验。
用ChatGPT分析定量数据：
T检验（T-test）：用于比较两组数据的均值差异是否显著。根据数据的配对情况，T检验可分为独立样本T检验和配对样本T检验。独立样本T检验用于比较两组独立样本的均值差异，而配对样本T检验用于比较同一组样本在不同条件下的均值差异。
方差分析（ANOVA）：用于比较三组及以上数据的均值差异是否显著。单因素方差分析用于比较一个因素下不同水平的均值差异，而多因素方差分析用于同时考虑多个因素对结果的影响。
卡方检验（Chi-square test）：用于分析分类变量之间的关联性，检验观察到的频数与期望频数之间的差异是否显著。在用户研究中，卡方检验常用于分析不同类别用户对某一特征的偏好差异。
相关分析（Correlation analysis）：用于衡量两个或多个变量之间的线性关系强度和方向。皮尔逊相关系数用于衡量连续变量之间的线性相关性，而斯皮尔曼等级相关系数用于衡量有序分类变量或非正态分布的连续变量之间的相关性。
回归分析（Regression analysis）：用于研究一个或多个自变量对因变量的影响程度。线性回归用于分析自变量与因变量之间的线性关系，而逻辑回归用于分析自变量对二分类因变量的影响。
主成分分析（PCA）：用于降维处理，将多个相关变量转化为少数几个不相关的主成分，以减少数据的复杂性。在用户研究中，PCA可用于识别影响用户行为的主要因素。
聚类分析（Cluster analysis）：用于将样本划分为若干个同质的子集，使得同一子集内的样本相似度高，而不同子集间的样本相似度低。在用户研究中，聚类分析可用于识别不同类型的用户群体。
方差齐性检验（Homogeneity of variance test）：用于检验不同组别数据的方差是否相等。方差齐性是进行T检验和方差分析的前提条件之一。
About Me
Chris Zhang, known as "科叔", is a seasoned user experience architect and agile coach with extensive experience serving both multinational corporations and local innovative enterprises. He previously worked as an engineer at Apple, where he gained a profound understanding of user-centered product development principles. After returning to China, Chris has been dedicated to providing consulting and internal training services to leading Chinese technology companies and industry unicorns, including ByteDance, Huawei, Alibaba, Baidu, vivo, Transsion, and Shokz. Chris excels at conveying professional and complex user research methods in engaging and practical ways, helping teams rapidly enhance their user research capabilities and practical skills. He is known for his ability to make product knowledge accessible and enjoyable, facilitating a deeper understanding of user experience design among his audiences.
Chris Zhang，昵称“科叔”，是一位资深用户体验架构师和敏捷教练，拥有丰富的跨国企业与本土创新企业服务经验。他曾在苹果公司任职工程师，深入理解以用户为中心的产品开发理念。回国后，科叔致力于为中国一线科技企业及行业独角兽提供咨询与内训服务，涵盖字节跳动、华为、阿里巴巴、百度、vivo、传音、韶音等知名企业。科叔擅长以有趣且实用的方式传递专业、复杂的用户研究方法，帮助团队快速提升用户研究能力与实践水平，促进对用户体验设计的深入理解。
Unlock hidden desires with AI tools, creating products that truly speak. 用AI工具解锁隐藏的欲望，创造真正触动用户的产品。
Comments