Enterprise Marketers Deploy AI Citizen Tools to Bypass Slow IT Cycles

企业营销人员部署AI平民化工具以绕过迟缓的IT周期 营销高管总结认为，赋能员工作为AI平民开发者能够消除工程积压，并将产品配方研发与软件交付周期从数周缩短至数小时。 来自媒体、房地产、制造业和风险投资领域的高管分享了将AI融入营销运营以提升速度、贴近度与效率的具体方法。讨论展示了多项实际应用落地案例，例如MeriCal的自动化配方引擎Nutravolta、eXp Realty在48小时内完成的定制界面部署与高管数字孪生，以及借助n8n和Julius AI等工具进行的数据编排。在整个讨论中，与会嘉宾主张将AI视为人类“思想领袖”的“思考伙伴”，并强调了员工技能提升、营销通才培养以及构建健全的企业合规护栏的重要性。 AI 原生的一人创业公司和精简团队正利用自主化工具套件，直接与传统跨国企业展开正面竞争。 (2:14) AI 在获客层面大幅压缩产品研发周期，MeriCal 的 Nutravolta 引擎可在短短一个下午内生成经过临床验证的配方与专利框架，颠覆了以往数周的研发流程。 (8:17) 客户触点正从传统静态应用向具备情商（EQ）的语音电商和生成式 3D 渲染流程演进，在加快创意速度的同时兼顾品牌治理。 (8:17) 企业级 LLM 与无代码工具（如 Lovable、n8n）相结合，让营销团队成为全民开发者，无需等待工程排期即可在 48 小时内搭建出定制化的内部工具及界面。 (16:06) 为缓解员工对岗位被替代的焦虑，企业应将人类定位为把控战略方向的“思想领袖”（thought leaders），而将 AI 视为执行辅助的“思考伙伴”（thought partners）。 (22:07)

Industry leaders argue that artificial intelligence drives customer acquisition and operational scale only when technical speed is anchored by strict compliance guardrails, internal citizen development, and human strategic oversight. Across diverse sectors, panelists emphasize that AI acts as an execution thought partner rather than an autonomous replacement for human leadership. Accelerating Customer Proximity and Commercial Velocity (8:17) Michael Berean of SiriusXM Media asserts that competitive advantage belongs to whoever gets closest to the customer through real-time account-based marketing, while using AI internally to eliminate process debt and surface high-intent buyers. In manufacturing, Josh Beaty highlights MeriCal's internal Nutravolta engine. Developed on a five-year repository of clinical research and regulatory label claim filings, Nutravolta synthesizes clinically validated probiotic formulations alongside patent blueprints in a single afternoon during client negotiations, replacing formulation timelines that historically took weeks. Consumer-facing deployments similarly target customer intimacy. Dimple Malkani of Solaris Venture Partners details Sephora's progression into voice commerce on smart home devices, where conversational models evaluate speech cadence and emotional sentiment to tailor product recommendations. Malkani also points to enterprise content generation partnerships, such as Coca-Cola leveraging NVIDIA and WPP to generate photorealistic 3D campaign assets, reconciling creative velocity with strict brand protection. Rapid Citizen Development and Cross-System Workflows (16:06) Marketing organizations are bypassing multi-month IT backlogs by using no-code platforms and foundational models to prototype internal solutions in under 48 hours. Wendy Forsythe shares that when Meta deprecated its Workplace platform, eXp Realty's 82,000 residential agents—who average over 55 years old and rely heavily on visual social feeds—resisted moving to Slack's channel-based layout. In response, eXp's marketing operations team used enterprise OpenAI models alongside no-code builder Lovable to construct a visual front-end skin over Slack in two days, resolving demographic user friction without traditional software development cycles. At the workflow level, Forsythe highlights the automation platform n8n as both a personal productivity tool and enterprise middleware. In real estate transactions, where closing operations often stall across disconnected software silos, n8n automates data handoffs between spreadsheets, CRMs, and email cadences, eliminating manual copy-pasting and human error during property closings. Workforce Evolution and Executive Digital Twins (20:20) To scale mentorship across 82,000 independent agents and 2,400 corporate staff, eXp Realty created digital twins of executive leadership using custom ChatGPT models informed by structured interviews. Staff across all tiers can asynchronously query these models for situational coaching, critique, and strategic guidance. Addressing employee anxieties regarding job obsolescence, leadership emphasizes that human workers remain the indispensable 'thought leaders,' while AI serves as an execution-oriented 'thought partner' incapable of originating strategy. Concurrently, marketing roles are pivoting from hyper-specialized profiles toward versatile generalists. Beaty explains that lean teams can master the foundational 20 percent of multiple disciplines and rely on AI tools for the remaining 80 percent of execution. Platforms like Julius AI allow non-analysts to ingest syndicated Circana or Nielsen datasets and identify market opportunities in minutes rather than days. Additionally, marketing teams deploy GPT assistants to review weekly task logs, identifying capacity bottlenecks and prioritizing high-return tasks without punitive oversight. Enterprise Governance and Rapid SaaS Obsolescence (37:54) Michael Berean warns that unlike startups with small operational blast radiuses, global enterprises face severe liabilities under regulations like GDPR. Major enterprise risks include employees uploading personally identifiable information into commercial large language models or deploying unsupervised browser agents directly into core environments like Salesforce. Rather than issuing blanket bans that stifle innovation, Berean recommends defining explicit data boundaries, educating business units, and utilizing AI tools to audit internal data architectures for processing errors. Panelists also observe that rapid advancements in foundational models and native AI browsing engines threaten traditional B2B SaaS categories with obsolescence. To manage tool adoption safely, organizations are formalizing internal experimentation by fully subsidizing enterprise licenses—including Gamma, Lovable, n8n, and Fixer.ai—and designating recurring calendar blocks, such as Thursday afternoon sandboxes, to foster innovation without overloading standard workloads. While the panel demonstrates that rapid prototyping and citizen development can dramatically compress operational timelines, enterprise adoption remains bounded by regulatory compliance, data privacy risks, and the uncertain longevity of third-party software vendors against fast-evolving foundation models.

在Ai4研讨会上，多位企业营销与技术运营主管指出，人工智能的核心价值在于缩短与客户的距离并重构内部敏捷性；企业应当将员工定位为主导战略的“思想领袖”，将AI作为执行层面的“思考伙伴”，并在严密的数据治理底线与全员赋能之间取得平衡。 压缩交付周期与深化客户连接 (8:17) SiriusXM Media的Michael Berean指出，现代营销的竞争壁垒在于谁能最大限度贴近客户，团队需借助AI实现实时个性化与基于客户的营销（ABM），精准对接决策者痛点并清理底层数据架构。在B2B制造领域，MeriCal的Josh Beaty分享了内部系统Nutravolta的实践：依托沉淀五年的临床试验和监管合规数据库，AI能在数分钟内生成具有科学验证依据的益生菌配方及专利蓝图。这使得原本需耗时数周的配方论证压缩至与客户谈判的当天下午完成，显著提升了品牌合作伙伴的上市速度。 在消费端与创意领域，Solaris Venture Partners的Dimple Malkani指出，AI正通过具备情绪智能（EQ）的人机交互深化品牌亲密度。例如Sephora利用智能家居语音模型分析用户的语调和情绪状态，提供定制化美妆建议；Coca-Cola则联合NVIDIA与WPP搭建实时3D创意资产管线，在满足创意团队快速迭代需求的同时，有效保障了跨国企业的品牌视觉规范。 全民开发者化解系统迁移与流程断层 (16:06) 面对技术栈变动与跨部门协作摩擦，营销团队正借助大语言模型演变为“公民开发者”。eXp World Holdings的Wendy Forsythe分享了一个案例：在Meta宣布下线Workplace平台后，该机构旗下8.2万名平均年龄超过55岁且习惯社交动态流界面的房产经纪人面临迁移至Slack的巨大阻力。营销运营团队结合OpenAI企业级模型与无代码工具Lovable，在48小时内为Slack定制开发出类社交动态的前端界面，绕过了长达数月的传统IT开发排期，消除了庞大经纪人团队的工具抵触情绪。 此外，Forsythe阐述了自动化编排工具n8n的应用实践。针对房产交易流程中各个孤立软件系统形成的信息孤岛，eXp不再依赖人工机械复制粘贴表格数据，而是构建跨应用工作流：自动抓取电子表格与部门更新，将其同步至CRM并触发定制跟进邮件。在复杂的房产交割阶段，这一自动化方案填补了老旧系统间的断层，降低了人为失误并加速了结算周期。 领导层数字分身与复合型通才崛起 (20:20) 为打破组织层级限制并普惠指导资源，企业开始利用ChatGPT创建管理层及创始人的“数字分身”。eXp通过结构化深度访谈，将高管的决策哲学、管理风格与长期愿景编码至定制模型中，无门槛开放给基层员工及独立经纪人，使其能够全天候异步获取决策反馈与战略解答。针对创意团队对职业替代的焦虑，Forsythe强调，人类员工必须是不可替代的“思想领袖”，而AI仅是“思考伙伴”，战略方向与业务判断始终需要人类经验把关。 在团队架构层面，传统深耕单一点位的专业分工正向“复合通才”转型。Josh Beaty指出，在仅由3人组成的精简营销团队中，成员需掌握跨文案、战略与数据分析的基础知识（前20%），并依靠AI完成剩余80%的执行。例如利用Julius AI平台，营销人员可在数分钟内直接解析Circana、SPINS或Nielsen等市场财团数据以挖掘空白市场，替代了以往数据分析师数天的Excel建模工作，大幅释放了管理者的精力并缓解会议疲劳。 严守数据合规红线与规范化沙盒机制 (37:54) 随着生成式工具在企业层面的普及，数据治理与风险管控成为规模化应用的关键前提。Michael Berean警告，相比初创团队，大型全球化企业承受着严苛的GDPR与隐私合规压力。员工若私自将包含个人身份信息（PII）的敏感表格上传至公共LLM窗口，或在Salesforce等核心生产系统中随意运行未受监督的自主浏览器Agent，极易引发合规灾难与数据泄漏雪崩。企业不应采取一刀切的全面禁令，而应联合法务部门制定清晰的合规红线，指导员工在安全边界内进行报表分析与系统排错。 为引导团队有序创新而非“野蛮生长”，eXp等企业推行了带薪且受保障的实验机制。管理层在每周四下午为全员划定专门的“沙盒时间”，并由公司全额出资采购企业级授权（包括用于快速开发的Lovable与n8n、用于创意排版的Canva Pro与Gamma，以及用于邮件智能分类的Fixer.ai），确保团队在不侵占日常业务考核的前提下，将回收的工时投入到高价值的流程重塑中。 小组讨论表明，生成式AI与自动化工具正在从根本上重塑客户触达与组织运营效率，但底层基础模型的快速迭代也威胁到诸多传统B2B SaaS软件的生存周期。如何在保护专有数据与合规安全的前提下，持续维护轻量级自主开发能力，仍是企业推进AI原生化不可回避的长期课题。

## Background 本内容由 Ai4 频道于 2025 年 8 月 31 日发布在 YouTube 平台上。视频记录了一场题为《AI 驱动的营销：精准且规模化获客》的圆桌研讨，汇聚了来自 ISS、SiriusXM、eXp Realty、MeriCal 及 Solaris Venture Partners 的多位高管与投资人，共同探讨企业如何利用 AI 预测消费者行为、优化广告支出并实现自动化获客与个性化运营。 ## TL;DR 圆桌嘉宾围绕 AI 在企业获客、组织转型与数据合规中的实战落地展开了深入交流。讨论涵盖了利用大模型打破粗放式泛化营销、MeriCal 借助自研 Neutravolta 系统实现临床配方的秒级生成，以及 eXp Realty 在 48 小时内利用 ChatGPT 和 Lovable 敏捷化解组织协作迁移危机等生动案例。与会者一致认为，人类员工应始终作为主导方向的“思想领袖”，将 AI 视为赋能思考的伙伴；同时强调大型企业必须建立清晰的安全与数据红线，并在组织内部营造沙盒文化，通过 n8n 等工具实现跨系统自动化以释放核心生产力。 ## Atmosphere & style 整场讨论节奏明快、务实且充满行业一线洞察，既有资深极客对拨号上网时代的幽默怀旧，也有对大模型快速演进引发初创企业生存危机的深切紧迫感。主持人与四位嘉宾互动默契，在“战略宏观构想”与“具体工具踩坑心得”之间自如切换。现场氛围既洋溢着技术解放生产力的兴奋，又贯穿着对合规治理与组织心理焦虑的严谨审视，带给观众一种在变革前沿既热血探索又保持清醒克制的实战质感。 ## Key takeaways - [08:17](https://www.youtube.com/watch?v=1sSJ27hyC84&t=497s) AI 在获客中的核心价值在于打破粗放式的泛化营销，通过精准的 ABM 与智能对话交互实现实时触达，并消除内部陈旧的数据与系统负债，从而优先识别高意向潜在客户。 - [11:23](https://www.youtube.com/watch?v=1sSJ27hyC84&t=683s) 实体制造与保健品代工巨头 MeriCal 借助自研系统 Neutravolta，将沉淀 5 年的海量临床研究数据库与 AI 结合，在数秒内自动生成合规配方并完成专利商标布局，将研发推向市场周期从数周大幅缩减至当天。 - [12:48](https://www.youtube.com/watch?v=1sSJ27hyC84&t=768s) 消费级 AI 正在向情绪智能（EQ AI）演进，如丝芙兰探索通过智能家居捕捉用户语调情绪提供护肤建议，可口可乐则联合英伟达与 WPP 借助生成式 AI 构建高品质 3D 品牌视觉资产。 - [16:06](https://www.youtube.com/watch?v=1sSJ27hyC84&t=966s) eXp Realty 营销团队在无技术开发介入的情况下，利用 OpenAI 企业级服务与 Lovable 仅耗时 48 小时便搭建出高视觉体验的前端，化解了 8.2 万名经纪人因习惯差异对 Slack 迁移的强烈抵触。 - [20:20](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1220s) 通过对高管进行深度结构化访谈，企业可将管理者的隐性经验提炼为全员公开可调用的“数字分身”（Digital Twins），打破层级信息壁垒，赋能日常方案评审与战略沟通。 - [23:16](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1396s) 缓解员工职业替代焦虑的关键在于确立“机器是思考伙伴，人类是思想领袖（Thought Leader）”的协作认知，引导团队聚焦策略方向与综合判断，而非陷入被替代的恐惧。 - [25:31](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1531s) 管理者可借助定制 GPT 收集团队日常任务与工作积压，智能评估任务优先级，其核心目的并非考核惩罚，而是解决“在制品（WIP）过多”并减轻管理疲劳。 - [28:28](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1708s) AI 时代推动了全能型/通用型营销人（Marketing Generalist）的回归，小团队可借助 Julius 等自然语言分析工具充当力量倍增器，几分钟内自动提炼复杂品类数据与市场空白点。 - [32:07](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1927s) Perplexity 打造 Comet 浏览器等突破性底层产品的极速演进，可能在短时间内淘汰大批 AI 原生初创公司，企业在采购 B2B SaaS 工具时需重点评估供应商的抗风险与长期生存计划。 - [35:18](https://www.youtube.com/watch?v=1sSJ27hyC84&t=2118s) 企业可通过设立每周四“沙盒时间”及全额采购 Lovable、n8n、Gamma、Fixer.ai 等工具账号，为员工提供无压力的实验环境，并通过自动化处理邮件与 PPT 为业务人员释放宝贵工时。 - [37:56](https://www.youtube.com/watch?v=1sSJ27hyC84&t=2276s) 大型企业在 AI 落地中必须将 GDPR 与数据合规置于首位，严防员工擅自将敏感 PII 上传至公开大模型，只有在清晰界定安全红线的前提下才能放心释放团队的创新能力。 - [42:08](https://www.youtube.com/watch?v=1sSJ27hyC84&t=2528s) 利用 n8n 工作流编排工具不仅能在个人层面串联表格、CRM 与自动化邮件触达，更可在核心业务层面作为桥梁自动同步多套遗留交易系统，彻底消灭人工跨系统搬运数据的低效痛点。 ## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=1sSJ27hyC84&t=4s) 圆桌开场：ISS 设施服务巨头与多元视角的 AI 探索 圆桌主持人由 ISS 北美区首席营销官（CMO）Marjorie McCarthy 担任。她首先向与会者介绍了 ISS 的业务规模与核心定位：作为一家全球领先的工作场所体验与设施服务公司，ISS 在全球拥有超过 32 万名员工，业务覆盖从《财富》100 强巨头到中端市场的各类企业，进驻场景涵盖现代化办公楼、机场及医院等核心场所。ISS 的服务范畴十分广泛，包括保洁、技术运维、餐饮配餐、前台与迎宾接待等，其核心理念在于为企业工作场所注入人性化的“款客服务”（Hospitality）。面对如此庞大的员工群体与多元的客户网络，如何借助 AI 将一线员工与企业客户更紧密地连接起来，成为了当前营销与运营的核心命题。 Marjorie 指出，本次圆桌汇聚了来自不同行业领域、代表不同企业规模与成长阶段的资深营销人与行业思想领袖，大家将从各自独特的业务背景出发，探讨 AI 在获客、运营与组织赋能中的前沿实践，并邀请现场嘉宾依次介绍各自的背景及所处企业的 AI 落地旅程。 --- ### [02:14](https://www.youtube.com/watch?v=1sSJ27hyC84&t=134s) 嘉宾介绍：从拨号时代的极客情怀到硬科技与 AI 原生创企投资 SiriusXM Media 的 B2B 营销运营（Marketing Ops）负责人 Michael Berean 率先发言。他在营销运营领域深耕超过 15 年，在加入 SiriusXM 之前曾自主创办并经营营销代理商，长期与众多全美及全球顶级品牌保持紧密合作。Michael 自称骨子里是一个“未来主义者”（Futurist）。在谈及 AI 时，他流露出极大的热情，并回忆起科技变革的历史脉络：“这就像上世纪 70 年代各种新发明问世，或是 80、90 年代互联网与拨号上网兴起时的浪潮一样。”他幽默地分享了一段童年往事——当年在用 America Online（AOL）拨号上网聊天时，只要母亲拿起家里的座机电话，网络就会被强行掐断，聊天也随即中断。Michael 强调，虽然当下身处前所未有的技术爆发期，但作为营销人必须保持理性平衡：绝不能“为了用 AI 而用 AI”，必须设立明确的业务目标，并在利用 AI 为团队创造更大业务影响力的同时，严格恪守对用户隐私（Privacy）、个人身份信息（PII）与数据合规的底线。 随后，Solaris Venture Partners 的风险合伙人 Dimple Malkani 进行了自我介绍。Solaris 主要专注于数据基础设施、AI、机器人技术（Robotics）以及人形机器人（Humanoids）等前沿硬科技领域的投资，尤其聚焦于推动亚洲创新技术落地美国的跨国项目。Dimple 是一名在硅谷 Web 1.0 时代就入局的资深科技营销与战略合作专家，曾在 NBCUniversal、辉瑞（Pfizer）和 AT&T 等跨国巨头工作。谈及当前的 AI 革命，她指出整个商业生态正发生剧烈分化：那些从诞生第一天起就以 AI 为底座的“AI 原生组织”（AI-native organizations）展现出了惊人的战斗力，传统大型品牌如今不得不正视并迎战那些仅由几个人甚至“一人团队”创立的初创公司所带来的颠覆性竞争。 --- ### [05:51](https://www.youtube.com/watch?v=1sSJ27hyC84&t=351s) 赋能生态与传统转型：房地产超级网络与 65 年益生菌制造巨头 eXp World Holdings 首席营销官 Wendy Forsythe 介绍了其母公司旗下的两大核心品牌：全球知名房地产经纪品牌 eXp Realty，以及拥有百年历史的个人发展与商业期刊品牌《SUCCESS》杂志（SUCCESS Enterprises）。其中，eXp Realty 的业务网络遍布全球 28 个国家和地区，为超过 82,000 名住宅房地产专业经纪人提供底层赋能支持。Wendy 形容团队目前已经“完全跌入了探索 AI 的兔子洞”（all the way down the rabbit hole）。在经纪业务端，他们将经纪人视为核心客户，帮助他们掌握并运用 AI 工具更好地服务终端买家与卖家；在企业内部，他们推动 2,400 名全职（W2）员工化身为“氛围编码”（vibe coding）与“全民开发者”（citizen developers），引导各岗位在日常工作流程中主动开发与集成 AI 工具，大幅提升运营效能。 MeriCal 营销总监 Josh Beaty 接着介绍了这家深耕实体制造的隐形巨头。MeriCal 位于加州奥兰治县（Orange County），是北美最大的益生菌与膳食补充剂代工商，拥有 65 年历史、600 余名员工以及 4 个制造基地，全美主流大型零售商、Vitamin Shoppe 和 GNC 货架上的众多补充剂产品均由其研发生产。Josh 坦言，虽然工厂制造环节高度自动化，但由于历史悠久，企业内部的协同与管理流程相对陈旧。在他加入公司的这一年半时间里，他受命牵头推动内部数字化转型，并率先在营销与 IT 部门之间搭建起 AI 探索试点，目前正处于快速推进的起步阶段。 - **ISS North America**：全球逾 32 万名员工，覆盖《财富》100 强及中端企业客户，涵盖清洁、餐饮、技术、前台等全方位设施与工作场所体验管理。 - **SiriusXM Media**：深耕 B2B 营销运营，聚焦精准获客、系统债务清理与数据合规。 - **Solaris Venture Partners**：专注数据基建、AI、机器人与人形机器人投资，聚焦亚洲到美国的前沿技术落地。 - **eXp World Holdings**：赋能全球 28 国逾 8.2 万名经纪人，内部 2,400 名员工积极实践“全民开发者”与 AI 工作流自动化。 - **MeriCal**：北美最大益生菌制造商，65 年历史、4 个生产基地，正从营销与 IT 端发起内部数字化与 AI 转型。 --- ### [08:17](https://www.youtube.com/watch?v=1sSJ27hyC84&t=497s) 获客核心逻辑：缩短客户距离与解决技术债务 主持人 Marjorie 将讨论引向本次圆桌的核心主题——客户获取（Customer Acquisition）。谈及如何利用 AI 建立与客户的深度连接，Michael Berean 开玩笑称团队正在利用 AI 谋求“统领全球”（global domination），随后严肃地阐述了他的营销哲学：在高度同质化与竞争激烈的市场中，“谁离客户最近，谁就能赢得胜利”（whoever is closest to the customer wins）。客户对品牌的忠诚度不是凭空产生的，促使客户信任并在产品生态系统中停留 5 到 10 年是一项极具挑战的任务。 Michael 认为，AI 对营销、销售及收入运营（Revenue Operations）最核心的价值在于： 1. **打破泛化营销**：告别过去粗放式的“一刀切”（Blanket statements）策略，充分认识到无论是 B2B 还是 B2C，决策链条的终点都是活生生的人（Human beings），必须借助 AI 迅速理解并分析其个性化需求。 2. **实时与精准触达**：在客户产生需求的瞬间，通过精准的账户营销（ABM）、智能对话交互体验（Chat experience）以及量身定制的解决方案，以极高的效率打动客户。 3. **消除内部系统与数据负债**：利用 AI 深度扫描内部业务系统，自动识别系统红线（Red flags）、冗余/膨胀数据（Bloated data）以及流程与技术债务（Process and technical debt）。企业规模越大，这些债务的滚雪球效应就越严重；借助 AI 自动化清理数据链路，营销团队能够精准识别并优先服务那些最接近成交阶段（Closer to an actual sale）的高意向客户，直接带动营收转化。 ```mermaid flowchart LR accTitle: AI-Driven Customer Acquisition Engine accDescr: Diagram illustrating internal system cleanup and external customer engagement powered by AI. subgraph InternalOps["内部运营与数据优化"] A["多源业务系统数据"] --> B["AI 内部诊断与清洗"] B --> C["消除技术/流程负债\n识别红线与膨胀数据"] end subgraph Acquisition["外部精准获客引擎"] C --> D["高意向销售机会优先排序"] D --> E["实时个性化互动\n(ABM / 智能会话)"] E --> F["缩短成交流程\n提升客户 LTV"] end ``` ### [11:23](https://www.youtube.com/watch?v=1sSJ27hyC84&t=683s) MeriCal 打造“Neutravolta”系统：临床数据与配方研发的秒级转化 MeriCal 市场营销总监 Josh Beaty 介绍了公司在研发领域的一项关键创新。过去五年间，MeriCal 的研发部门一直在系统化构建一个庞大的临床研究与科学报告数据库。由于营养与保健品行业对产品标签上的功效宣称（Claims）有着极其严苛的合规与监管要求，所有面向客户的产品配方都必须具备经过验证的临床数据支撑。为了将这套沉淀了五年的海量数据真正转化为生产力，MeriCal 自主开发了一套内部代号为“Neutravolta”的智能系统。 ```mermaid flowchart TD accTitle: MeriCal Neutravolta 研发流程图 subgraph DataEngine["数据与智能引擎"] A["5年沉淀的临床研究与学术报告数据库"] --> B["Neutravolta 智能配方系统"] B --> C["合规与法律功效宣称引擎"] end subgraph Output["业务落地与交付"] C --> D["全新合规配方生成"] D --> E["商标与专利自动化布局"] D --> F["当日高管会议即时交付"] end ``` 通过 Neutravolta 系统，团队只需轻点几下鼠标，即可在几秒钟内生成一套完全经过临床验证、且此前从未存在过的全新配方，并同步启动商标和专利布局。Josh Beaty 强调，这彻底改变了公司的推向市场速度（Speed to Market）：在与重要客户召开高规格决策会议时，早晨提出的配方设想，下午团队就能直接拿出完备方案并告知对方“配方已就绪，所有临床文献与佐证数据均已打包”，而这类工作在过去通常需要耗费数周的研发周期。 - **合规核心**：产品功效标签的合法性完全依赖于严格的临床试验证明，数据库与 AI 的结合从源头解决了法规审核风险。 - **商业响应力**：将配方研发、合规论证到专利规划的周期从“数周”压缩至“当天下午”。 --- ### [12:48](https://www.youtube.com/watch?v=1sSJ27hyC84&t=768s) 消费级 AI 范式：Sephora 的情感化交互与 Coca-Cola 的 3D 内容重塑 Solaris Venture Partners 的风险合伙人 Dimple Malkani 随后从跨行业投资视角分享了 B2C 领域的代表性 AI 应用。首先是长期走在数字化前沿的美妆零售巨头丝芙兰（Sephora）。丝芙兰此前已在官方 App 中普及了基于人脸识别的虚拟试妆功能，帮助用户在海量 SKU 中挑选发彩、护肤品及彩妆。而丝芙兰眼下正在探索更具突破性的方向——赋予 AI 情绪感知能力（EQ/Emotive sensing），并将其深度整合至 Alexa 和 Google Home 等智能家居生态中。 ```mermaid flowchart LR accTitle: 消费级 AI 创新架构 subgraph Sephora["Sephora 智能家居情绪交互"] S1["家庭语音设备 (Alexa/Google Home)"] --> S2["语调与情绪识别引擎"] S2 --> S3["个性化护肤建议与语音下单"] end subgraph CocaCola["Coca-Cola 3D 视觉生成"] C1["可口可乐品牌团队"] --> C2["WPP 营销集团 + NVIDIA"] C2 --> C3["高质量 3D 品牌资产与实时内容"] end ``` 用户未来不仅能直接通过语音向家庭设备咨询专业护肤建议，AI 还能敏锐捕捉用户说话时的语调（Tone）与潜在情绪，理解当下真实的肌肤状态与情感诉求，进而直接完成语音下单。另一个典型案例是传统消费品巨头可口可乐（Coca-Cola）与英伟达（NVIDIA）以及全球广告巨头 WPP 的战略合作。可口可乐利用生成式 AI 工具进行 3D 内容图像的实时构建。这一实践不仅打破了“AI 将杀死广告代理商”的行业预言，也化解了传统品牌团队在采用 AI 工具时对“内容品质失真”的保守顾虑，在保持高标准品牌资产的同时实现了内容生产力的飞跃。 - **情绪智能（EQ AI）**：AI 正在从单纯的文字/图像识别进化到声音语调与情绪状态的实时分析，重构家庭场景下的消费决策。 - **代理商角色的演变**：顶尖代理机构通过与底层算力与芯片巨头（如 NVIDIA）联合，以高品质 3D 生成技术重塑自身价值链。 --- ### [16:06](https://www.youtube.com/watch?v=1sSJ27hyC84&t=966s) eXp Realty 的 48 小时破局：利用 ChatGPT 与 Lovable 化解 Slack 迁移危机 eXp Realty 首席营销官 Wendy Forsythe 结合具体运营场景，分享了 AI 如何在 48 小时内化解一场席卷全公司的组织协作危机。eXp Realty 作为拥有 82,000 名房地产经纪人的大型上市企业，旗下经纪人社群的人口画像高度偏向于 55 岁以上、以女性为主且高度依赖视觉呈现的群体。过去，这些经纪人主要依托 Meta 旗下的 Workplace 进行房源推介、图片展示和内部引荐业务。然而，随着 Meta 宣布关闭 Workplace，公司被迫全员迁移至 Slack。 这一调整遭到了经纪人群体的强烈抵触，习惯了社交化、视觉化界面的房产经纪人极度不适应 Slack 偏向文本流的极客风格。恰逢 OpenAI 发布新一代模型体系，eXp Realty 的营销团队在没有借助任何技术开发部门、没有走冗长跨部门需求文档（PRD）评审流程的情况下，直接通过 OpenAI 企业级服务结合 Lovable 工具，在不到 48 小时内快速搭建出一套兼具 Workplace 视觉体验与交互习惯的界面，底层则直接对接 Slack 的通信接口。 ```mermaid flowchart TD accTitle: eXp Realty 48小时快速解决方案流程 subgraph Problem["组织痛点与冲突"] P1["Meta 关闭 Workplace"] --> P2["82,000 名 55+ 视觉型经纪人抵触 Slack"] end subgraph Solution["48小时敏捷交付"] P2 --> S1["OpenAI 企业级 API + Lovable"] S1 --> S2["构建高视觉体验的前端交互层"] S2 --> S3["无缝打通 Slack 底层通讯生态"] end subgraph Impact["业务价值"] S3 --> I1["免去数月研发周期与 PRD 流程"] S3 --> I2["团队彻底摆脱传统流程束缚"] end ``` 这次快速迭代实现了双赢局面：经纪人重获直观的图文交互体验，而公司则成功将全员沉淀在统一的 Slack 生态内。Wendy 激动地表示，这是半年前团队完全无法想象的自由度与生产力：“我们真正解除了团队身上的束缚（taken the handcuffs off），只要扫清体制障碍，你就会发现组织内极具天赋的员工能创造出令人惊叹的成果。” - **用户画像断层**：55 岁以上的传统房产经纪人对生产力工具（Slack）有天然排斥，高度依赖社群化与图像化沟通。 - **去中心化敏捷开发**：营销团队摆脱传统 IT 研发管线，通过 AI 编排工具在两天内独立完成企业级 UI/UX 改造。 --- ### [20:20](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1220s) 组织全员的“数字分身”：打破层级的高管 AI 教练体系 除了解决协作工具危机，Wendy Forsythe 还分享了 eXp Realty 在组织治理层面的另一项深度探索：为全公司管理者打造“数字分身”（Digital Twins）。团队利用 ChatGPT 对各级管理人员进行了系统化、结构化的深度访谈，提取其决策逻辑、管理哲学与沟通风格，从而为每位管理者生成一个随时待命的 AI 导师。 ```mermaid flowchart LR accTitle: 数字分身赋能体系架构 subgraph Input["知识沉淀"] A["管理者与创始人"] --"深度结构化访谈"--> B["ChatGPT 知识与逻辑建模"] end subgraph Output["全员赋能场景"] B --> C["'如果是 Wendy 会怎么做？'<br/>日常业务决策与方案评审"] B --> D["基层员工直连 CEO 分身<br/>咨询公司 2030 战略愿景"] end ``` 在日常工作中，团队成员可以随时向 Wendy 的数字分身提问：“如果是 Wendy，她会如何处理眼前的情况？”或者请求分身对某个方案提供评审与辅导。更为关键的是，这套数字分身系统在组织内部是完全公开透明的——从新入职的基层员工到一线房产经纪人，任何人都可以直接访问公司创始人兼 CEO 的数字分身，并向其抛出如“公司在 2030 年的核心战略规划是什么？”等宏观问题。这种全公开机制不仅重塑了跨层级沟通的透明度，也为作为上市公司的 eXp Realty 探索组织效率边界开辟了全新路径。 会议主持 Marjorie McCarthy 随后顺势将话题引向深层次的组织变革：在全员 AI 转型的过程中，领导者应当如何有效引导团队的心态演变，同时又该如何与企业核心利益相关者及高层战略对齐，以确保 AI 探索能持续转化为切实的商业价值。 - **知识资产化**：通过大模型访谈将管理者的隐性经验转化为全公司随时可调用的显性资产。 - **组织扁平化与透明度**：基层员工与一线销售可以直接对话企业一把手的数字分身，打破信息孤岛。 ### [23:16](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1396s) 克服员工恐惧与思维转变：AI 是思考伙伴，你是思想领袖 当讨论到团队对 AI 的焦虑时，Wendy Forsythe 指出，营销团队内部普遍存在关于“我的工作会不会消失”的恐惧，尤其是设计师等特定岗位。她提到自己大概每隔一个月就会和设计师进行一对一谈话，面对设计师提出的交流请求，必须展开坦诚的对话。Wendy 鼓励团队不能从恐惧的角度出发，而必须从兴奋的角度看待变化，因为技术变革不会消失，恐惧只会限制未来的可能性。她鼓励员工通过参与更多项目、寻找能够为业务做贡献的新领域来拓宽视野、扩展技能集。 针对这种心理状态，团队内部确立了一个核心原则：“机器是你的思考伙伴，而你是思想领袖”（The robot is your thought partner. You are the thought leader）。Wendy 强调，只要员工始终保持思想领袖的定位，将 AI 作为辅助思考的伙伴来应对每一个问题、挑战和机遇，就能在变革中找到笃定与安全感。每当团队出现不安或不适感时，管理层都会引导大家回归到这种思维模式中。 ```mermaid flowchart LR accTitle: 人类与 AI 的协作角色分工 subgraph 人类角色 A["人类员工<br/>（思想领袖 Thought Leader）"] end subgraph AI 角色 B["AI 工具 / 机器人<br/>（思考伙伴 Thought Partner）"] end A -->|"主导方向、提出策略、综合判断"| C["解决业务问题与抓住机遇"] B -->|"提供灵感、数据处理、执行支持"| C ``` --- ### [25:31](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1531s) 新旧技术共存与利用 GPT 解决管理疲劳 Michael Berean 赞同 Wendy 的观点，并从技术发展史的角度阐述了技术演进中的“共存”现象。他以 iPhone 为例指出，尽管 iPhone 是有史以来最伟大的发明之一，能够高效处理日常 80% 的事务，但人们依然离不开笔记本和台式电脑来处理剩余 20% 的复杂工作，新旧技术实现了长期共存。相比之下，有些曾被大肆炒作的概念（例如元宇宙 Metaverse）最终却如过眼云烟，沦为小众爱好者的玩具。虽然某些传统体验会像曾经在高中深夜去 Tower Records 唱片店买唱片配 Wendy's 汉堡那样，最终变成被技术取代的怀旧记忆，但在 AI 演进过程中，人机共存仍是主旋律。 针对现场众多带团队的管理者，Michael 谈到了管理者普遍面临的“Zoom 疲劳”与“会议疲劳”，以及身兼“队员与教练”（Player-Coach）双重角色的精力撕扯。他分享了一个自己正在测试且不涉及任何 PII（个人身份信息）和隐私风险的内部实践： - **任务与效率汇报**：让团队成员直接向定制的 GPT 汇报每周完成的任务以及顺延到下周的工作事项。 - **智能分析与优先级排序**：利用 GPT 评估工作效率，分析哪些任务最重要、哪些可以延后、哪些应当优先处理。 - **减轻在制品负荷**：该举措的目的并非为了进行绩效考核或问责惩罚，而是帮助团队解决“在制品（Work in Progress）过多”的困境，使有限的资源集中在能够带来更直接、更快速回报的高价值营销事项上，并向高层领导清晰汇报。 --- ### [28:28](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1708s) AI 时代“全栈/通用型营销人”的重新崛起 Josh Beaty 探讨了 AI 如何重塑营销岗位画像，特别强调了“通用型营销人”（Marketing Generalist）的回归。对于资源精简的小团队而言，这种趋势尤为显著。Josh 所在的 MeriCal 拥有 600 名员工，但营销团队只有 3 个人：作为负责人的 Josh 本人、一名设计师以及一名与销售团队紧密配合的品类分析师（Category Analyst）。在缺乏人力编制的情况下，领导者必须将 AI 构建为自身的“力量倍增器”（Force Multiplier）。 Josh 分享了利用 AI 赋能团队的具体案例与职业经历思考： - **自动化品类数据洞察**：分析师原本需要耗费大量时间从 Circana、SPINS、Nielsen 等行业数据源拉取数据并在 Excel 中编写复杂公式。现在团队借助名为 Julius 的工具，设定好业务边界规则后，只需用自然语言提问寻找市场空白点（White Space），工具在几分钟内即可输出原本需数天人工分析才能得出的产品机会报告。 - **从 T 型人才转向全能通用型**：十年前行业崇尚在单一领域钻研极深的“T 型营销人”（T-shaped Marketer）。Josh 毕业初期作为社交媒体协调员，曾打造出无需付费广告便拥有百万粉丝的 Facebook 页面。但随着职业发展，领导者需要掌握 20% 的全方位综合技能，其余 80% 的具体执行则交由 AI 工具或团队完成。 - **跨界复合背景的求职优势**：在当前的就业竞争中，拥有全盘把控能力的通用型人才往往更具优势。Josh 透露自己在应聘现职时，雇主放弃了单一领域拔尖的文案或专才，正是因为他“样样都懂一点，足以形成实战杀伤力”（had enough of everything to be dangerous）。 --- ### [32:07](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1927s) 技术迭代极速演进下的行业冲击与初创企业危机 针对 AI 落地速度与组织转型节奏的议题，Dimple Malkani 结合自己作为 Perplexity AI Fellow 的一线观察，分享了对极速技术演进的深刻担忧。 由于较早接触到了 Perplexity 的前沿能力，并亲眼目睹了诸如 Comet 浏览器与 Labs 等产品的早期内部演示，Dimple 直言自己产生了一种“胃部下沉的揪心感”（a pit in my stomach）。她指出，Perplexity 仅用了大约 7 个月的时间就开发出了 Comet 浏览器，而此类突破性底层产品的集中发布，可能会在瞬间颠覆并淘汰大批 AI 原生初创公司以及现有的 B2B 软件企业。这种前所未有的产品迭代与颠覆速度，正在给整个技术生态和各类组织带来巨大的生存危机与适应挑战。 ### [33:42](https://www.youtube.com/watch?v=1sSJ27hyC84&t=2022s) 评估供应商生命力与告别键盘和 PPT 的未来工具 作为企业决策者，在采购 AI 工具或尝试成为早期采用者时，必须保持审慎。当前市面上充斥着大量可能在短期内倒闭的 B2B SaaS 工具，因此与供应商洽谈时，首要任务是查明他们的“长久生存计划”（longevity plan），搞清楚他们究竟要如何与行业巨头抗衡。 在谈及 AI 工具与生产力提升时，一些前瞻性的创新正在颠覆日常工作流。例如 WhisperFlow，它很可能会彻底淘汰人类沿用已久的键盘输入方式：未来无论是撰写 Slack 消息还是处理工作邮件，一切都可以通过语音直接输入完成。另一个颠覆性工具是 Gamma（通过 Perplexity AI 研究项目了解到），它直接革新了制作幻灯片的方式。大家过去在每张 PPT 页面上耗费的大量排版与制作时间将被一扫而空，过去人们“希望 PowerPoint 彻底消失”的梦想正借助 AI 成为现实。密切关注这些工具的前沿动态，是企业保持敏锐度与竞争力的关键。 - 采购 SaaS 与 AI 软件时，必须考察初创供应商面对科技巨头时的抗风险与长期生存能力。 - WhisperFlow 致力于通过语音输入替代键盘，直接打通邮件与团队即时通讯。 - Gamma 正在颠覆传统的演示文稿制作流程，大幅减少在单张 PPT 幻灯片上耗费的人工工时。 --- ### [35:18](https://www.youtube.com/watch?v=1sSJ27hyC84&t=2118s) 打造 AI 采纳文化：沙盒时间、全员付费订阅与收件箱救星 要在团队内部实现 AI 技术的快速采纳，仅有工具是不够的，核心在于建立相匹配的企业文化。很多员工在面对新技术时常伴有焦虑：“我怎么能在忙完本职工作的同时，还要去学透这些复杂的 AI 工具？”为了打破这一顾虑，管理层必须主动为员工腾出学习的时间与空间。 团队确立了每周四下午的固定机制：周四下午被设置为专属的“沙盒时间”（sandbox time），员工可以在此期间自由探索、实验 AI 工具。同时，公司承担所有软件成本，为每位员工统一采购企业级账号，包括 Lovable、n8n 工作流自动化账号以及 Canva Pro 企业版，杜绝员工自掏腰包的现象。此外，针对让创作者抓狂的演示文稿，团队引入了 Gamma；而面对最令人头疼的电子邮件堆积问题，团队引入了 Fixer.ai。Fixer.ai 能够自动接管收件箱，撰写回复草稿，并将邮件自动梳理至 8 个不同分类中，每周为员工和房产经纪人省下数小时的宝贵时间，让他们将精力释放到真正的创意与业务增长中。 ```mermaid flowchart TD accTitle: 企业 AI 采纳赋能闭环 subgraph Culture["企业文化与时间保障"] A["每周四下午：沙盒探索时间"] --> B["消除学习焦虑，主动释放精力"] end subgraph Tooling["企业买单的全套生产力矩阵"] C["Lovable & n8n: 自动化工作流"] D["Gamma & Canva Pro: 演示与视觉设计"] E["Fixer.ai: 智能分类收件箱与代写回信"] end subgraph Outcome["业务成效"] F["每周节省数小时日常杂务"] --> G["投入核心创造力与高阶业务扩展"] end Culture --> Tooling Tooling --> Outcome ``` - **文化保障**：设立周四下午“沙盒时间”，让员工在带薪时间内无压力学习与实操 AI 工具。 - **全额报销工具链**：企业为员工统一采购 Lovable、n8n、Canva Pro 及 Gamma 等账号。 - **邮件提效解决方案（Fixer.ai）**：能够自动分类邮件至 8 个类目并代写回复，直接为团队和代理人每周“赎回”数小时工时。 --- ### [37:56](https://www.youtube.com/watch?v=1sSJ27hyC84&t=2276s) 大企业的 AI 治理：合规风控、数据安全与划定创新边界 初创公司或中小型企业在应用 AI 时试错成本相对较低，能以极快的速度推进；但对于大型跨国企业而言，AI 治理、合规性与 GDPR（通用数据保护条例）监管则是首要防线。过去 24 个月 AI 的演进速度堪称惊人，但如果缺乏监控和制度规范，技术落地极易失控。 企业绝不能放任员工“各自为政（going rogue）”——比如擅自将包含 PII（个人身份信息）的表格复制粘贴进公共 ChatGPT，或者随意在浏览器中让未经授权的 AI Agent 接入 Salesforce 等核心业务系统。面对高管层与董事会，最重要的汇报要点是“制衡机制是否健全”以及“是否规避了公司暴露在合规风险之下”。这种风险一旦缺乏管控就会像滚雪球一样演变成雪崩式的严重数据泄露事故。管理的真谛在于清晰划定“可做与不可做”的边界：只有当团队明确知道红线在哪里，才能在安全合规的围栏内充分施展创意，利用 AI 进行数据分析、生成洞察摘要、优化广告投放，甚至监控系统漏洞与陈旧数据。 - **大企业与初创公司的差异**：初创团队轻装上阵，而全球化大型组织必须把 GDPR、数据主权与治理框架置于首位。 - **严禁无序使用**：严禁员工擅自将敏感客户数据（PII）上传至公开大模型，严防未经验证的 AI Agent 侵入核心 CRM 系统。 - **划定红线以释放创意**：先确立安全边界，团队才能放心利用 AI 开展异常排查、数据提炼及广告策略分析。 --- ### [40:54](https://www.youtube.com/watch?v=1sSJ27hyC84&t=2454s) 快问快答：嘉宾最爱的 AI 潮流与工具 在进入观众提问前，主持人和嘉宾们展开了一轮简短的“快问快答”（Quickfire），各自盘点近期最令自己着迷的 AI 工具或趣味趋势： | 嘉宾 / 角色 | 偏好的 AI 趋势 / 工具 | 核心看点与使用评价 | |:--- |:--- |:--- | | **Michael Berean** | AI 虚构乐队（Fake AI Band） | 互联网上根本找不到该乐队真身，却在各大流媒体音乐平台狂揽海量播放量，被评价为一个极其搞笑且有趣的社会学实验。 | | **Dimple Malkani** | 机器人格斗（Humanoid Boxing） | 关注旧金山车库里人形机器人的对战与拳击实验，展现了强烈的赛博未来感。 | | **Wendy Forsythe** | Fixer.ai | 实用主义选择，彻底解决日常收件箱灾难，是不可或缺的邮件神器。 | | **Josh Beaty** | Julius AI (julius.ai) | 专为非技术背景营销人打造的利器，原本需要耗费数小时的专业数据分析，几分钟内即可搞定，令人直呼“仿佛置身天堂”。 | --- ### [42:08](https://www.youtube.com/watch?v=1sSJ27hyC84&t=2528s) 现场问答：利用 n8n 打造跨系统自动化与房产交易流 在问答环节，现场观众针对工作流自动化平台 n8n 发问，探讨其最具价值的 Agentic（智能体）工作流实践。作为一款开源与企业级兼具的编排工具，团队不仅鼓励全员在个人日常工作中构建自动化，还将其深度嵌入到核心业务主干中。 在个人及轻量应用层面，团队成员最常用 n8n 解决碎片化数据流转痛点。例如，自动从电子表格中提取信息，清洗后同步推送到 CRM 客户关系管理系统中，再触发个性化邮件发送；或者接收来自特定业务部门的数据，整合并归档至表格及其他内部系统中。而在企业核心业务层面，作为一家涉及大量复杂流程的房地产企业，每笔交易的达成通常要在多个互不连通的业务系统之间来回倒手数据。团队借助 n8n 填补了系统之间的流转断层，实现了数据在交易系统间全自动平滑同步，彻底根除了人工录入和手动跨系统搬运数据的低效与失误。 ```mermaid flowchart LR accTitle: n8n 跨系统数据自动化流转架构 subgraph Input["数据源与触发端"] A["业务部门提交 / 电子表格"] end subgraph Engine["n8n 自动化与编排中枢"] B["数据清洗与格式转换"] C["交易系统断层对接逻辑"] B --> C end subgraph Output["业务终点系统"] D["企业级 CRM"] E["自动化通知 / 邮件触达"] F["房产交易管理系统"] end A --> Engine C --> D C --> E C --> F ``` - **个人轻量级工作流**：将外部表格、部门需求、内部 CRM 及后续跟进邮件自动串联，摆脱手工复制粘贴。 - **房产交易主业务协同**：在交易终审与交割过程中，n8n 充当桥梁，填补多套遗留系统与管理平台之间的功能空隙。 - **信息流自动化理念**：只要存在跨系统的数据流动，一律利用 n8n 实现端到端自动化，消灭人工搬运。 ## Notable quotes > [02:14](https://www.youtube.com/watch?v=1sSJ27hyC84&t=134s) 这就像上世纪 70 年代各种新发明问世，或是 80、90 年代互联网与拨号上网兴起时的浪潮一样。 — Michael Berean > [02:14](https://www.youtube.com/watch?v=1sSJ27hyC84&t=134s) 绝不能“为了用 AI 而用 AI”，必须设立明确的业务目标 — Michael Berean > [08:17](https://www.youtube.com/watch?v=1sSJ27hyC84&t=497s) whoever is closest to the customer wins — Michael Berean > [11:23](https://www.youtube.com/watch?v=1sSJ27hyC84&t=683s) 配方已就绪，所有临床文献与佐证数据均已打包 — Josh Beaty > [16:06](https://www.youtube.com/watch?v=1sSJ27hyC84&t=966s) 我们真正解除了团队身上的束缚（taken the handcuffs off），只要扫清体制障碍，你就会发现组织内极具天赋的员工能创造出令人惊叹的成果。 — Wendy Forsythe > [20:20](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1220s) 如果是 Wendy，她会如何处理眼前的情况？ > [23:16](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1396s) 机器是你的思考伙伴，而你是思想领袖（The robot is your thought partner. You are the thought leader） — Wendy Forsythe > [28:28](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1708s) 样样都懂一点，足以形成实战杀伤力（had enough of everything to be dangerous） — Josh Beaty > [32:07](https://www.youtube.com/watch?v=1sSJ27hyC84&t=1927s) a pit in my stomach — Dimple Malkani > [35:18](https://www.youtube.com/watch?v=1sSJ27hyC84&t=2118s) 我怎么能在忙完本职工作的同时，还要去学透这些复杂的 AI 工具？ ## Mentioned Marjorie McCarthy, ISS North America, Michael Berean, SiriusXM Media, America Online (AOL), Dimple Malkani, Solaris Venture Partners, NBCUniversal, 辉瑞 (Pfizer), AT&T, Wendy Forsythe, eXp World Holdings, eXp Realty, 《SUCCESS》杂志 (SUCCESS Enterprises), Josh Beaty, MeriCal, Vitamin Shoppe, GNC, Neutravolta, 丝芙兰 (Sephora), Alexa, Google Home, 可口可乐 (Coca-Cola), 英伟达 (NVIDIA), WPP, Meta, Workplace, Slack, OpenAI, Lovable, ChatGPT, iPhone, Tower Records, Circana, SPINS, Nielsen, Julius AI, Perplexity AI, Comet, WhisperFlow, Gamma, n8n, Canva Pro, Fixer.ai, Salesforce ## Video description Marjorie McCarthy, Chief Marketing Officer, ISS North America Wendy Forsythe, Chief Marketing Officer, eXp Realty Michael Berean, Director B2B Marketing Operations, SiriusXM Josh Beaty, Director of Corporate Marketing, MeriCal Dimple Malkani, Venture Partner, Solaris Venture Partners Abstract This session will focus on how AI is transforming the way businesses acquire new customers, enabling more precise targeting, personalized outreach, and scalable campaign strategies. Panelists will discuss how AI-driven tools can analyze customer behavior, predict purchasing patterns, and automate marketing efforts to reach the right audience at the right time. Learn how companies are using AI to optimize ad spend, generate high-quality leads, and create personalized customer journeys that convert prospects into loyal customers. The discussion will also cover best practices for integrating AI into your customer acquisition strategy and the challenges of balancing automation with a human touch. [Source](https://www.youtube.com/watch?v=1sSJ27hyC84)

[2:14] Michael Berean: 这就像上世纪 70 年代各种新发明问世，或是 80、90 年代互联网与拨号上网兴起时的浪潮一样。 [2:14] Michael Berean: 绝不能“为了用 AI 而用 AI”，必须设立明确的业务目标 [8:17] Michael Berean: whoever is closest to the customer wins [11:23] Josh Beaty: 配方已就绪，所有临床文献与佐证数据均已打包 [16:06] Wendy Forsythe: 我们真正解除了团队身上的束缚（taken the handcuffs off），只要扫清体制障碍，你就会发现组织内极具天赋的员工能创造出令人惊叹的成果。 [20:20] 如果是 Wendy，她会如何处理眼前的情况？ [23:16] Wendy Forsythe: 机器是你的思考伙伴，而你是思想领袖（The robot is your thought partner. You are the thought leader） [28:28] Josh Beaty: 样样都懂一点，足以形成实战杀伤力（had enough of everything to be dangerous） [32:07] Dimple Malkani: a pit in my stomach [35:18] 我怎么能在忙完本职工作的同时，还要去学透这些复杂的 AI 工具？

[2:14] Dimple Malkani: AI-native organizations are a force to be reckoned with [2:14] Wendy Forsythe: all the way down the rabbit hole [20:20] Wendy Forsythe: What would Wendy do? [22:07] Wendy Forsythe: The robot is your thought partner. You are the thought leader. And as long as you always come from a position of being the thought leader and the robot being your thought partner, you shouldn't be fearful. [37:23] the gift of AI [37:54] Michael Berean: going rogue [37:54] Michael Berean: an avalanche problem [40:54] Marjorie McCarthy: I will say what I mentioned earlier, Julius AI, just because I'm a creative marketer, not a technical analyst. So when I saw what I could do in minutes, I'm like, 'Oh my gosh, like this just saved me like hours in my analyst hour.' I was like, 'I'm in heaven.'... I don't represent it. I just learned it last week. I've only used it three times, but I'm already mind-blown.

[0:04] Hi everyone. I'm Marjorie McCarthy. I'm [0:07] chief marketing officer of ISS North [0:10] America. If you don't know us, we are a [0:14] global workplace experience and facility [0:17] services company. We have over 320,000 [0:21] employees around the world. We are the [0:24] people that power the buildings that we [0:27] all work and have experiences in. So we [0:31] do cleaning, we do technical services, [0:34] we do culinary services, we do workplace [0:37] and re reception services, we are [0:40] focused on delivering hospitality in the [0:43] workplace. And because of our large [0:46] workforce and because we are working [0:49] with customers across the Fortune 100 [0:52] down to the mid-market and we're in [0:54] offices, we're in airports, we're in [0:57] hospitals. [0:59] how we leverage our workforce and how we [1:04] make sure that we're connecting our [1:05] workforce to our customers is critical [1:08] for us and how AI connects into that [1:12] conversation especially how how it [1:14] connects to our customers is is really [1:16] important for us. So I feel really [1:18] excited to to get to be a part of this [1:21] panel today and we have a really amazing [1:26] group of marketers and industry thought [1:28] leaders here and what's really exciting [1:31] is that this is a very diverse group. [1:34] We've got different industries [1:35] represented. We have different company [1:38] sizes and scales and growth journeys [1:41] represented. [1:43] And so I just we're really going to have [1:45] an interesting conversation. So to kick [1:49] us off, [1:51] if y'all could each introduce [1:52] yourselves, tell us a little bit about [1:55] where you are, uh who you're with, but [1:58] also tell us a little bit about where [2:02] you are in your AI journey. So, you [2:05] know, as a marketer, as a marketing [2:07] organization, [2:09] um as an enterprise, um maybe we can [2:11] just kick it off there and and Michael, [2:12] we'll start with you. [2:14] >> Hey everybody, Michael Barian. I head up [2:16] B2B marketing ops at SiriusXM Media. [2:20] I've been uh in the marketing ops game [2:22] for well over 15 years. Um before being [2:27] at Sirius, kind of ran an agency as [2:28] well. So, kind of got that experience [2:30] working, you know, very closely with [2:32] some larger companies. you know, [2:34] national and global brands. I I think at [2:38] heart, uh, I'm a futurist, you know, so [2:40] I think that, you know, there's a lot of [2:42] varying opinions on AI, uh, application [2:46] of AI, like we all we all have [2:48] different, um, kind of takes on it. But, [2:52] you know, I think right now for me, like [2:54] I'm just fascinated by it. Um, and [2:56] that's not just cuz I'm a marketer and [2:57] cuz I work in a company that's, you [2:59] know, owns technology companies as well, [3:01] but um, just the ability to [3:05] engage with it and to live in this era [3:07] that we're in, I think is really [3:08] special. And, you know, maybe people [3:10] said that was uh, the same thing, you [3:12] know, in the 70s when, you know, we had [3:14] different inventions come out then and [3:16] then the 80s and 90s and the internet [3:18] and dialup and, um, but I don't know if [3:20] if everyone's as old as I am. Um, but [3:24] you know, I remember having to get [3:25] kicked off of America Online when my mom [3:28] needed to use the phone, you know, and [3:30] uh my my chat getting disrupted. So, um, [3:33] I'm excited, you know, personally. And, [3:35] you know, where I think I balance that [3:38] with is someone said it earlier in [3:41] another keynote is having goals, right? [3:43] We don't just apply AI to apply AI. We [3:47] want to have goals and we want to [3:48] respect privacy and personally [3:50] identifiable information as marketers [3:52] and data. Um, but we also want to be [3:55] optimistic and use it to create more uh [3:58] impact for our teams. [4:03] >> Thank you. My name is Dimple Malcani and [4:06] I am currently a venture partner at [4:09] Solaris Ventures and what we do at [4:13] Solaris is invest in projects related to [4:16] data infrastructure, AI, robotics, [4:19] humanoids with a specific focus on Asia [4:22] into the US. And you know to answer [4:25] Marjgerie's question um my whole career [4:28] has sort of come together to to do what [4:31] I'm doing today. I'm a career marketer [4:34] and partnerships person uh major brands [4:38] you know NBC Universal, Fizer, AT&T uh [4:41] but at the same time I'm a Silicon [4:43] Valley baby a native and started my [4:45] career very early in the Bay um from web [4:49] 1.0 onwards. So I've been in continuous [4:52] practice of taking major brands and [4:55] taking the latest technology and merging [4:57] them. And so it's super exciting being [4:59] part of the AI revolution and being able [5:01] again to take all that and culminate it [5:03] into what I'm doing today. And you know [5:06] my thoughts on AI is you know we have to [5:08] embrace it. Uh [5:11] just early adoption is wonderful. What [5:13] we're seeing with a lot of our um [5:15] startups that we work with is that the [5:19] ai native organizations are a force to [5:22] be reckoned with. you know, traditional [5:24] and um major brands are having to [5:28] compete with what you know, a one-man [5:30] show startup is doing. And um you know, [5:33] we all know examples of that. And so, uh [5:37] what v what we're doing at Solaris is [5:39] even looking past that in terms of [5:41] physical hardware, AI, humanoids, and um [5:45] you know, a lot of trends that we'll [5:47] speak about that are coming up. So, [5:49] thank you. [5:51] >> Good afternoon everyone. I'm Wendy [5:53] Foresight. I'm the chief marketing [5:54] officer at EXP World Holdings. You would [5:57] know uh perhaps two of our main brands. [6:00] EXP Realy uh EXP Realy operates in 28 [6:05] different countries around the world. We [6:07] power um just over 82,000 residential [6:11] real estate professionals in their [6:14] brands. Um, as well we operate Success [6:17] Magazine through Success Enterprises, [6:20] which is just over a hundredyear-old [6:22] personal development brand. You've [6:24] probably seen the magazine um, if you [6:26] travel through an airport. Um, and [6:28] through both of those brands, we are, I [6:31] like to say, all the way down the rabbit [6:33] hole of using AI to empower both our [6:38] customers who on the uh brokerage side, [6:41] which is where I spend most of my time, [6:43] um, empowering our customers, who we [6:46] view as being our agents, our [6:48] residential real estate agents, to use [6:50] AI in their business to um, help them [6:53] serve their customers, which would be [6:55] their buyers and sellers out in the [6:57] marketplace. place and help empower our [6:59] employees. So, we have about 2400 W2 [7:03] employees that uh help serve our agents. [7:06] Um, and we're very much uh vibe coding [7:10] citizen developers finding ways um to [7:13] help our employees leverage AI in their [7:17] day-to-day workflows. So, we'll we'll be [7:20] talking a lot more about that. [7:25] >> Good afternoon. Uh my name is Josh [7:27] Batty. I'm director of marketing at [7:29] Maracal uh out in Nor County, [7:31] California. Um we are the largest [7:34] probiotic manufacturer in North America. [7:35] So you see supplements and big retailers [7:39] and vitamin shop and GNC. It's a good [7:41] chance we probably develop those in the [7:43] house. Um we have a company about 600 [7:45] people in four locations. And so we've [7:48] been around for about 65 years as a [7:50] company. While manufacturing has a lot [7:52] of automated processes, internal [7:54] processes are sometimes a little [7:56] archaic, especially when you have a [7:58] legacy company that's been doing certain [7:59] things a certain way. So, in my time at [8:02] Marcal within the last year and a half, [8:04] I've kind of been charged with [8:05] championing how to really innovate our [8:08] digital kind of transformation [8:10] internally, starting with the marketing [8:12] and IT department. And so, we're in our [8:14] early stage, but I'm excited to see what [8:15] we're doing. So, yeah. [8:17] >> Awesome. Thank you. So this c uh this [8:21] panel is about customer acquisition. So [8:24] why don't we start there? Let's start [8:25] with the customer. [8:28] You all have some really great examples [8:31] of of how you've used AI to create [8:34] connection um and an experience with [8:37] your customers. And so maybe we can get [8:39] into some real life examples. Um [8:43] Michael, I'm going to go to you first [8:44] because you had some some really [8:46] interesting points there. [8:49] Yeah. So, we're we're really focused on [8:51] using AI for global domination, right? [8:54] That's uh [8:56] all jokes aside, um I think as [8:59] marketers, we understand that, you know, [9:01] whoever's closest to the customer wins, [9:03] right? We're creating brands. We're [9:04] creating, you know, energy with a [9:06] customer um that gets them to, you know, [9:09] be a recurring customer or trust us or, [9:12] you know, be in our our product [9:13] ecosystem for 5 10 years, right? That's [9:16] a big ask, you know, of anyone today. I [9:18] mean, there's so many competitors out [9:21] there that our customers could go to, [9:23] right? So I think when we look at AI, [9:26] you know, really where I see it being so [9:28] impactful for marketers um and people [9:32] working in sales, marketing and revenue [9:33] ops is how do we use the AI AI to create [9:37] a closer connection to be more real time [9:40] to get to them when they need us more to [9:43] create more personalization whether [9:45] that's through ABM or it's through a [9:47] chat experience or it's through just an [9:50] experience that works best for that [9:52] specific individual. I think in the past [9:55] we tried to take like blanket statements [9:57] or blanket concepts and apply them to [9:59] all customers but we're talking about [10:00] human beings at the end of the day right [10:02] no matter what B2B B TOC it doesn't [10:04] really matter people and humans are at [10:07] the end of the decision-m right now when [10:08] it comes to products and solutions so [10:11] using AI to like better understand their [10:13] needs right to uh to analyze their needs [10:16] faster to understand them quicker and to [10:19] get them to a product or a solution that [10:22] works for them in a in a quicker way, in [10:25] a more efficient way. I think that's [10:27] what puts us ahead of our own [10:28] competitors, right? When we're looking [10:29] at our own competitive market. Um, [10:32] that's really where I'm just excited and [10:34] focused about that, right? Is filling [10:35] data gaps, um, personalization, ABM, um, [10:39] like I said, real-time engagement. And [10:41] then the other thing we've been doing [10:43] that I've been looking at, you know, [10:45] closely is how do we use AI to analyze [10:48] our own internal systems for, you know, [10:51] errors, red flags, you know, uh, bloated [10:54] data. Um, you know, we see a lot, you [10:56] know, both process and technical debt, [10:59] um, that we need to clean up. I think [11:01] everybody's company, you know, deals [11:03] with that. And then the bigger your [11:04] company, the more that compounds. So [11:06] again, that allows us to prioritize [11:09] customers who are closer to an actual [11:10] sale, you know, and have more impact on [11:13] that that revenue when we're using AI. [11:16] >> Josh, you also had a really interesting [11:19] example of how AI is impacting your [11:23] customer experience. You want to share [11:25] that? [11:26] >> Yeah. So, we've had a lot we have a lot [11:28] of interesting things that we're doing [11:29] at Marel right now. So for the past I [11:32] want to say five years in our research [11:35] and development department we've develop [11:36] we've been developing a database of [11:40] research reports clinical studies [11:42] because everything we want to make for [11:44] our customers we want it to be [11:45] clinically backed and substantiated and [11:47] that comes with a lot of things like [11:48] regulatory. You want to be able to have [11:50] a claim on a label you have to [11:51] substantiate that for it to be legal. [11:53] And so over the past 5 years, this [11:56] database has been created, but we had to [12:00] find a way to really make use of it [12:01] effectively. And so we created this new [12:04] system called Neutravolta. It's kind of [12:05] our own little internal name. And with a [12:08] click of a few buttons, we can spin up [12:10] any kind of formula that's clinically [12:12] substantiated. We can create formulas [12:13] that have never been created. And then [12:15] we can create a trademark, patent, all [12:18] kinds of things around that. And so [12:19] that's something that customers really [12:22] like when we can come speed to market. [12:24] When you have a big stakes meeting, you [12:26] can go back that afternoon be like, "We [12:28] have a formula." People are like, "Wait, [12:29] what? This takes weeks to develop." It's [12:31] like, "Well, we have the team that [12:32] already developed the the whole [12:34] foundation, but we can come to you with [12:35] a formula that says it's ready to go. So [12:38] it's substantiated. Here's all the [12:39] research. If your people want to read [12:41] it, it's ready to go." [12:43] >> During our next time together, we're [12:44] going to have you demo that for us [12:46] because it sounds really cool. Um, [12:48] Dimple, you have the perspective of [12:52] across many industries and maybe you [12:56] could share a couple of examples just [12:58] from the broader industry of of where [13:01] you've seen AI really impacting um, our [13:04] ability to access customers impactfully. [13:08] >> Yeah, thank you Marjorie. Um, and I [13:10] think a couple great examples to give [13:11] will be in the BTOC realm, uh, with [13:14] brands we're all familiar with and [13:16] actually piggy piggybacking off of what [13:18] Michael just said. So, one of my [13:20] favorite examples and brands that I [13:22] think does a great job of being an early [13:24] adopter of technology is Sephora. And [13:27] Sephora, you know, has already in the [13:30] last few years adopted on their app for [13:33] someone to take a photo of themselves [13:35] and try on the various SKUs that Sephora [13:39] has. And if any of you have walked into [13:41] a Sephora store, you know, there's, you [13:43] know, unlimited amount of colors and [13:45] shades and products from hair care to [13:47] skincare to makeup. And so that's [13:50] something that they've already been [13:52] doing. But what they're looking to do [13:54] down the road, um, and this is something [13:57] that is really neat that we're also [13:59] seeing in robotics is sort of this EQ or [14:02] this emotive um, sense that AIS are [14:07] starting to train on. And so what [14:10] Sephora will be doing is integrating [14:12] into Alexa and Google Homes so that you [14:15] can get skincare advice right from your [14:19] um home um device as well as at some [14:24] point they will be able to pick up the [14:26] tone and of you of of the person [14:29] speaking. So pick up my tone to [14:31] understand what my needs are. And I [14:33] think that's super cool. And so they're [14:35] already working on a way to sort of put [14:37] your Sephora order in through these home [14:40] devices using your voice and picking up [14:43] your emotions. So that's pretty darn [14:45] cool. Um, another example that I really [14:48] like is taking a really old traditional [14:50] brand that we all know, Coca-Cola. And [14:53] so Coca-Cola recently partnered with [14:55] Nvidia and WPP uh one of the major [14:59] agencies on 3D uh images for their [15:03] content and you know so a lot of us in [15:06] the AI world is definitely in the [15:08] marketing world thought that agencies [15:10] would die with the invent of AI but you [15:13] know agencies are smart and they've [15:15] adopted the AI tools and so they are you [15:18] know right there partnering with Nvidia [15:21] and bringing that instant um quality [15:25] brand imaging that Coca-Cola consumers [15:28] and brand loyalists are used to seeing [15:31] because something I've also heard are [15:33] you know when brand teams are trying to [15:35] adopt AI tools and um are questioning [15:40] the quality right they've been putting [15:41] out quality content and all of a sudden [15:44] they're using an AI tool and they're [15:46] really slow to adopt it or being very [15:47] conservative and then there's tension [15:50] between sort of product or product [15:52] marketing and brand and uh so a brand [15:55] like Coca-Cola has solved that problem [15:57] by still using the agency and the latest [16:00] tech companies to produce very high [16:03] quality content. [16:06] >> Very cool. Thank you. Okay, we're going [16:09] to turn a bit now to talking about [16:12] marketing operations because one of the [16:15] the conversations we had before this [16:18] panel is how can we be leveraging AI in [16:22] our marketing operations [16:24] to free up or economize resources so [16:28] that we can spend more time focusing on [16:30] the customer. Wendy, you you have done [16:34] this really impactfully in your [16:36] organization. So tell us more about [16:38] that. [16:40] >> Yeah. Um as as I mentioned, we've been [16:42] all the way in on AI across all of our [16:45] departments in our company, but [16:47] especially in marketing. So um we have [16:50] an enterprise account with Open AI. And [16:53] I'll give you an example that is sort of [16:55] hot off the presses in the last 48 [16:58] hours. So, um, we anxiously sort of had [17:02] a watch party on Monday as Chat GBT5 was [17:06] unveiled to the world. Um, anybody [17:08] playing with it yet? [17:10] Super cool. We're all kinds of excited [17:13] about it. Um, one of the challenges that [17:17] um, real world problem we have. So, I [17:19] mentioned our customers are our real [17:20] estate professionals. We have 82,000 of [17:23] them. um they their primary form of [17:26] communication with one another um within [17:30] our ecosystem has been um Meta's um [17:35] workplace and of course Meta announced [17:37] that they were retiring workplace and [17:40] the demographic of the average real [17:42] estate professional is sort of a 55 plus [17:47] year old individual here in the US skews [17:51] heavily female they love Facebook Right? [17:54] So, they've loved workplace and they're [17:56] they're sending referrals to one [17:58] another. They're posting their listings. [18:00] They're posting photos. They're heavily [18:02] visual humans. So, we've migrated our [18:06] entire organization over to Slack. And [18:09] this is just internal like EXP agentto [18:12] agent communication with one another. [18:15] You can imagine they're hating Slack, [18:18] right? We didn't make this decision. [18:21] This was a meta decision. Slack is the [18:23] best option. They hate it. Since Monday, [18:27] we've been able to stand up with new [18:31] sort of um chat GBD5 [18:34] and because of the enhancements that [18:36] OpenAI made with conjunction with now [18:40] because it's playing so much better with [18:42] lovable. So through sort of the hooks [18:44] that we can now get through chatgbt5, [18:48] lovable and slack almost like a look and [18:53] feel of workplace that will satisfy the [18:57] visual element that they were missing [18:59] still keep them connected into Slack [19:02] which is keeps us happy because we want [19:04] them in our ecosystem but gives them [19:07] that user visual experience that they [19:09] were missing like in under 48 hours. So [19:13] it's providing a winwin [19:15] and that's the kind of stuff that I [19:17] didn't need my dev team. I didn't need, [19:20] you know, a bunch of like like I didn't [19:23] need to go through all the stuff that I [19:25] would have had to go through, right, [19:27] with like crossf functional like docs [19:30] and requirements and back and forth and [19:33] months of like project stuff like we [19:36] just spun it up and like that is the [19:40] freedom and the power that AI has given [19:43] a marketing team that like we couldn't [19:45] even have dreamed of six months ago. [19:48] when we started this. Um, and like we [19:52] we've literally just taken the handcuffs [19:54] off of our team and and there's some [19:57] pain points with that, but the [20:00] creativity and the power it has given [20:03] everyone in our organization is just [20:06] amazing. You got to remove the [20:08] roadblocks and you will find some [20:11] amazingly talented people in your [20:14] company that will do things that will [20:16] blow your mind. Before [20:20] we move on, do do you want to give your [20:22] I thought your digital your digital twin [20:25] example was was really interesting. [20:27] >> So, another thing that one is just so [20:28] hot off the presses that like it's just [20:30] so cool. Um, so another thing that we [20:33] did um we created digital twins um of [20:38] our leaders. So um anyone that's sort of [20:41] managing a team in our company has [20:44] created working working with chat GBT a [20:47] digital twin of themselves. So we worked [20:50] with chat uh we had chat sort of [20:52] interview us to essentially provide a [20:57] coach that's accessible for all of our [21:00] staff to [21:03] talk to us anytime. So, if you know you [21:06] had a question like, "What would Wendy [21:08] do?" Any of my staff can talk to my [21:11] digital twin um and ask me that question [21:14] or how does Wendy feel about this or [21:18] coach me on this situation or provide [21:21] critique on this? Um they can work with [21:24] the digital twin to coach um through [21:27] that. Um and and that goes all the way [21:30] up the organization. our CEO, our [21:32] founder has a digital twin and the [21:35] digital twins are open so that everybody [21:37] in the organization has access to [21:39] everyone else's digital twin. So [21:42] everyone from an entrylevel employee to [21:46] any of our real estate agents can access [21:48] our founder and CEO's digital twin and [21:52] ask questions like where do you see our [21:54] company in 2030? And we're a publicly [21:56] traded company. So these are, you know, [21:59] interesting things to navigate through. [22:02] Um, which is a whole other panel [22:04] discussion. [22:07] Um well taking a lead from that though [22:10] we we talked a lot about um in our in [22:13] our pre- discussion about leading teams [22:18] whether it's a marketing organization or [22:20] otherwise through AI transformation [22:25] and how how are you all and I'd love for [22:29] a couple folks to to respond to this. [22:32] How are you all thinking about leading [22:35] your teams through AI transformation? [22:39] Um, you know, and all the fields that [22:41] that come along with that potentially, [22:44] right? And then in addition to leading [22:47] teams is a two-part question. How are [22:50] you thinking about working with your [22:53] stakeholders and your leadership to [22:58] to position [23:01] your AI journey in a way that that [23:05] drives business impact and and that [23:07] ensures that that connectivity with with [23:10] your leadership and with your [23:11] stakeholders. [23:14] I'll give that up to anybody or Wendy if [23:16] you want to kick off or Michael or [23:18] >> Yeah. [23:19] So it's a lot of communication there. [23:21] There is fear especially from the [23:23] employees and I'll speak from sort of [23:24] the marketing team specifically right [23:27] because the marketing team hears a lot [23:29] about you know are is my job going to be [23:32] gone right so especially specific parts [23:35] of the marketing team like my designers [23:38] my designers are fearful right and and I [23:40] have like one-on- ones with my designers [23:43] probably every other month right I've [23:44] got a designer that says like can I have [23:46] some time um and and it takes some frank [23:49] conversation So those conversations for [23:52] me is like look you've got to look for [23:54] ways to provide value and you've got to [23:56] get excited about this because it [23:58] doesn't h you can't operate from a place [24:00] of fear. You've got to operate from a [24:03] place of excitement because this is not [24:06] going to go away and your fear is only [24:10] going to inhibit the possibilities. So [24:13] let's look for the possibilities. And [24:16] what I've been encouraging all of the [24:18] team is to look for those possibilities. [24:22] So that means get involved in more [24:24] projects, look for more um areas of the [24:28] business where you can contribute. So it [24:31] is really about sort of expanding your [24:34] horizons and looking to like grow that [24:37] sort of skill set. Um but it's you know [24:42] it there is a a mindset that we talk a [24:45] lot about is that the the robot is your [24:49] thought partner. You are the thought [24:51] leader. And as long as you always come [24:54] from a position of being the thought [24:55] leader and the robot being your thought [24:58] partner, you shouldn't be fearful [25:00] because the the robot can't be the [25:02] thought leader. They can be your thought [25:04] partner. And if you leverage the robot [25:06] as your thought partner and stay strong [25:09] in you're the thought leader, bringing [25:11] the thought leadership to every problem, [25:14] challenge, and opportunity, stay strong [25:17] in that and find comfort in that. And so [25:20] we bring it back to that sort of um [25:23] mindset whenever sort of discomfort [25:26] comes into the conversation. [25:31] >> Yeah, I think Wendy brings up really [25:32] good points. Um, I I think you know [25:35] something to pay attention to as well is [25:38] like the iPhone is arguably one of the [25:41] biggest inventions ever, but yet we [25:44] still work on laptops and we still work [25:47] on desktop computers sometimes, right? [25:49] Um, because we found that even though [25:52] it's efficient for 80% maybe of the [25:55] things we do, this 20% it's not, right? [25:59] I need to sit down. I need to use this [26:00] laptop. I mean, so I think that was a [26:02] really interesting moment where we saw [26:04] this coexistence happened where the new [26:06] technology did not replace the old [26:09] technology, right? Um, and it's happened [26:11] a few times and then some things come [26:13] along that we think are going to be a [26:14] big hit and and they're just they're [26:17] just vapor, you know, like remember the [26:18] metaverse, remember how big that was [26:20] going to be, you know? Um, it came and [26:22] it went and it's a hobbyist thing now. [26:24] So, I think um as we move forward [26:27] through this AI kind of evolution, I [26:30] really do feel like it's going to be a [26:32] coexistence. You know, I feel like, you [26:34] know, some things are going to get [26:35] replaced the way Tower Records got [26:37] replaced, which by the way, I used to [26:39] love going to Tower Records in high [26:40] school late at night, right? You know, [26:42] um get a Junior Bacon Cheeseburger from [26:44] Wendy's, hit up Tower Records, you know, [26:46] and that's a nostalgic experience now [26:48] for me. And I think it'll be very [26:50] similar. you know, we'll have these kind [26:51] of nostalgic experiences that get [26:54] overtaken by the tech, but some things [26:56] will remain. Um, you know, just in this [26:59] room, show of hands, how many people [27:01] manage other people, though? I want to [27:02] know just like how many people have [27:03] direct reports? Okay, it's a lot. It's a [27:05] good good amount. [27:07] I don't know about you guys, but you [27:08] know, I get Zoom fatigue. You know, I [27:11] get meeting fatigue. You know, there's [27:13] days when I'm like, just give me 5 hours [27:15] of like uninterrupted time to just do [27:18] what I need to do. And when you're in [27:19] that player coach role for a company, [27:22] right, I think AI can be extremely [27:24] powerful. You know, one of the things [27:26] I've been testing like just personally [27:28] right now that involves no PII, it [27:30] involves no real, you know, risk is [27:33] having people report into a GPT, right? [27:36] that can then analyze our tasks, what [27:38] we've accomplished for that week, what [27:40] we're going to be, you know, postponing [27:42] till next week, and starting to gauge [27:44] efficiency, you know, and just letting [27:46] us know like what's important, what's [27:48] what can be pushed, what we should be [27:49] prioritizing. Not because we want to get [27:53] performance reviews or people are going [27:56] to get, you know, in trouble, but just [27:58] because I think all of us in this room [27:59] are probably struggling with too much [28:01] work in progress, right? there's just so [28:03] much to get done and there's only so [28:05] many resources we have and I think when [28:07] you leverage AI like that that reports [28:10] up to leadership very well right when [28:12] you can when you can create more [28:14] efficiency for your team and as [28:16] marketers prioritize things that are [28:18] going to have a more direct and faster [28:20] return so that's that's what I've seen [28:24] >> thank you um okay so [28:28] looking a little bit to the future [28:31] obviously there are many considerations, [28:33] many trends that were we all as [28:36] marketers and and leaders are thinking [28:39] about. Each of you all had some really [28:42] interesting insights on some some of [28:45] those specific trends leading into this [28:48] conversation. [28:49] So, I'm hoping that we can take a few [28:51] minutes to just tick through some of [28:53] those. And um Josh, I'm going to start [28:55] with you. um you raised a really [28:58] interesting point about the the rise [29:01] again of the marketing generalist thanks [29:04] to AI. You want to talk a little bit [29:05] about that? [29:06] >> Sure. I guess I'll come from this the [29:08] point of if you're in a smaller team [29:12] this is going to be really good for you. [29:14] Larger teams it can be a different [29:16] discussion perhaps but my team is very [29:18] leanal. We have 600 employees but we [29:21] have a marketing team of three. It's me [29:23] the head of marketing. I have a designer [29:24] and I have a category analyst that works [29:27] closely with our sales team. And so when [29:30] you can really scale yourself using AI, [29:34] it you're you you need to be a force [29:36] multiplier somehow someway. So if you [29:38] don't have the headcount to help you be [29:40] a force multiplier as a leader, you got [29:42] to build the solutions around you. And [29:44] so using the different tools and [29:47] resources and those kind of things help [29:49] scale. Um, one way that I help scale our [29:53] anal an analyst work is, you know, she [29:57] spends a lot of time pulling data and [29:59] pulling the reports from various um kind [30:02] of internal sources like Circana, Spins, [30:04] Neielson, things like that where there's [30:07] a tool, I'm not sponsoring it, but [30:10] called Julius where you can instead of [30:12] spending all the time running [30:14] formulations in Excel, you can ask for [30:17] to find the white space in certain areas [30:18] to find a product opportunity with a [30:20] with a client. And so you put it you [30:23] upload your your info into that and you [30:26] put the guardrails around it and it can [30:28] spit out all kinds of interesting things [30:30] that might have taken a few days manual [30:32] work as an analyst but it pops it on in [30:35] just a few minutes and I'm kind of blown [30:37] away at how you can scale yourself. So, [30:41] also with that said, I think about a [30:44] decade ago, the whole idea was being a [30:46] TJ marketer. And that's what I was kind [30:48] of told when I first started right out [30:49] of college was go deep, you know, go [30:52] deep in something. Mine was social [30:53] media. I started as a social media [30:55] coordinator, right out of grad school. [30:57] And so, I got really good at that. I [30:59] built a Facebook page of a million [31:00] customers with zero paid ads. You can't [31:02] do that anymore. Um, you know, that was [31:04] like my bread and butter back then. But [31:06] I found over the years that as I grew, [31:08] my skill set had to expand and I had to [31:10] be able to do 20% of everything [31:13] as a leader, but then allow the tools [31:15] and resources or people on your team to [31:18] do the rest of the 80%. And so I think [31:21] now if you are a generalist, it's going [31:24] to be your in your advantage to get that [31:26] job. You know, if you're it comes down [31:29] to three candidates, probably the [31:32] generalist is going to get that. That's [31:33] how I got my job. I learned just a few [31:35] months ago because I was asking my boss, [31:37] I'm like, "So, what what made you chose [31:38] me out of everyone?" Well, we had this [31:40] really good copywriter. We had this [31:41] really good person that was good at [31:42] this, but you just had enough of [31:44] everything to be dangerous. And so, that [31:47] can win you jobs sometimes. Might have [31:48] been different, you know, for years ago, [31:50] but I think the generalist is perhaps [31:51] the way to go. It's not the the fancy [31:54] cool kind of thing, but it's that's [31:56] where the value is. And then when you [31:58] come in with that, you can coach, you [31:59] can do all kinds of things, and you can [32:01] be a force multiplier. And that's where [32:02] you can, I think, really make the money [32:03] going forward if you're kind of [32:05] ambitious in that sense. [32:07] >> Yeah. Thank you. Um, so another one of [32:10] the considerations we talked about is is [32:16] rapid adoption, [32:18] navigating the speed of transformation. [32:21] Um, and just how were how were each as [32:25] marketers and in in our organizations [32:28] thinking about [32:30] the speed at which we're adopting and [32:32] and navigating that? And and Wendy and [32:34] Dimple, you both had really interesting [32:36] insights on on that particular topic. [32:41] And I will I will let y'all choose who [32:43] goes first there. [32:46] >> Thank you. Yes, I you know this came up [32:49] when we were um throwing ideas around [32:51] and what I'm noticing so I'm also an AI [32:55] fellow for Perplexity. So I've been [32:58] exposed to what Perplexity can do, you [33:01] know, fairly early. example being Comet [33:04] and Labs and seeing demos of that very [33:08] early you know my first sort of um I got [33:12] you know a pit in my stomach going there [33:14] are all these founders you know whether [33:16] the AI native companies or or otherwise [33:19] where you know a release of like the [33:21] comet browser is going to um kill a lot [33:26] of those startups and lot many B2B [33:29] software companies um in one go I mean [33:32] it took perplexity seven months to build [33:34] comet I believe and it's going to take [33:37] out a lot of companies and so I worry [33:40] for you know organizations that are [33:42] buying some of these AI tools and are [33:45] trying to be early adopters um or are [33:47] using you know some B2B SAS that are [33:50] that you know are going to be gone so I [33:53] think as um business leaders we need to [33:56] think about uh what are the right tools [33:59] and when you are talking to vendors [34:01] figure out what's their longevity plan [34:02] and how are they competing against um [34:05] some of the big guys out there. So [34:07] that's um that's something I I just [34:09] wanted to touch on one other thing which [34:11] is in terms of AI tools um and [34:14] efficiencies [34:16] uh whisper flow for example I think [34:20] would be something you all should look [34:22] into um which basically without it's [34:25] going to take away the keyboard uh which [34:27] is a technology we have been using for a [34:29] long time but you'll be able to speak [34:31] everything um from your Slack messages [34:34] to your emails and things like that. So [34:36] in terms of futuristic [34:38] um efficiency and and where AI is going [34:40] with spiritual and gamma gamma is [34:42] another one and these I've learned [34:43] through the AI perplexity um fellowship. [34:46] It's not something I'm promoting per se [34:48] but just from my insights um gamma will [34:52] take out PowerPoint and how much time we [34:55] spend on each slide um will go away. And [34:58] so I think these are the type of things [35:00] that are going to I mean did we ever [35:01] think PowerPoint was going to go away? I [35:02] know a lot of us dreamed of PowerPoint [35:04] going away. Um, but I think that it's [35:06] finally here. And so when we talk about [35:08] what's coming in the future and what to [35:10] be aware of, like that's that's what I, [35:13] you know, keep my finger on that uh to [35:16] stay relevant. [35:18] >> The the rapid adoption, if you want your [35:21] team to really lean into it, you've got [35:24] to give them the tools and then you've [35:25] got to create a culture um for it. So [35:29] one of the things is, you know, if [35:31] you're going to create that culture, [35:33] you've got to give them time and space [35:35] to learn. So we heard, I guess, a fear [35:39] early on and and I say early on like [35:42] last year, right? Like we're literally [35:45] like months into all of this of like how [35:47] do I do my job and learn all this AI [35:50] stuff, too? So we immediately said, [35:52] okay, we've got to make space for people [35:54] to do that. So in our team, we we [35:57] started like Thursday afternoon. Um we [36:00] said like Thursday afternoon you block [36:01] off like that's your like sandbox time, [36:04] right? Like that's when you play and and [36:07] everybody has an enterprise account. [36:11] Everyone has a lovable account. [36:13] Everybody has an N8. We do a lot of [36:15] workflow so they all have their own NAN [36:17] accounts. We're paying for all of this [36:19] for them. So they're not having to like [36:21] pay like out of their own pocket for any [36:23] of this because they're using it to [36:25] build [36:27] efficiencies in their jobs. So they have [36:29] more time to be creative. We also tried [36:33] to solve for them pain points. So we use [36:36] Gamma because you can spend a lot of [36:38] time as a creative building [36:40] presentations. We have enterprise Canva [36:42] Pro accounts for everyone because that's [36:44] another efficiency especially on the [36:46] marketing team for them. But email, [36:49] email is a lovehate relationship. So we [36:52] wanted to find a solve for email. So we [36:55] found fixer.ai. [36:57] So Fixer is an AI tool that will save [37:01] hours every week in your inbox. AI is [37:04] your whole inbox. Writes your replies [37:07] for you. S like sorts emails into eight [37:10] different categories. Go get Fixer. I [37:12] don't sell Fixer. I'm not a resale. It's [37:14] a third party. But Fixer like is such a [37:18] gift to your inbox. So we got our whole [37:21] staff, we we provide it to all of our [37:23] agents now because it gives you back [37:26] hours every week. So when you get those [37:29] hours back, then you can go leverage [37:31] them to do other things. So we made that [37:35] commitment to our staff was like if [37:37] we're going to like really take our [37:40] business to the next level with the gift [37:43] of AI, we'll help you find the time and [37:46] space so that we can go then create with [37:50] it other efficiencies. So those are just [37:53] some examples of things that we've done. [37:54] >> Yeah, that's really interesting. Thank [37:56] you. Um, and then Michael, you shared a [38:00] passion for education [38:03] on just compliance and thinking about [38:07] GDPR and just all of the different [38:10] considerations around governance. Um, [38:13] when we think about leveraging AI tools. [38:15] So, talk to us a little bit about that. [38:18] >> Yeah, I think that's the big difference [38:19] between a smaller company and a larger [38:21] company or especially a global company. [38:23] you know, startups can move really fast, [38:26] you know, um, not just startups, [38:27] startups, medium-sized businesses, they [38:29] have they have, I think, a little less [38:31] risk to mitigate, right? Um, I think, [38:36] you know, one thing, look, I like I said [38:38] earlier, I'm I'm a futurist. Like, I [38:40] love all of this stuff. You know, every [38:42] day I'm I'm learning more and um I'm [38:45] just amazed by where things are going. [38:47] And imagine where things are going to be [38:49] in five or 10 years. It's like [38:51] mind-blowing how fast the last 24 months [38:53] have moved. Um, but but with that said, [38:57] I think we also need to balance that [38:59] with an understanding that if we don't [39:02] monitor things, if we're not careful, [39:04] things can get out of control, right? [39:06] you need to have that kind of mindset as [39:07] well where you can't just have people [39:10] going rogue um you know putting [39:13] spreadsheets into GPT of personally [39:16] identifiable information or you know [39:19] unleashing agents into a Salesforce [39:21] instance or something you know in the [39:22] browser and so it's really I think [39:25] education is a big part and I think [39:27] that's what speaks up to senior [39:30] leadership and seuite executives is that [39:32] that type of thinking is in place where [39:34] there is checks and balances first and [39:36] we're not creating exposure for the [39:39] company and I'm not like obviously an [39:41] attorney. Um, but I've worked with [39:43] enough compliance and you know legal [39:46] people in my career to understand that [39:49] like you know this is a risk that if you [39:51] don't get on top of it it just kind of [39:54] again will start to snowball until you [39:56] have an avalanche problem you know and [39:58] you have a data leak and you have you [40:00] know um a true issue. So I think it's [40:02] really about educating your teams, you [40:04] know, even in your small business units [40:06] or the teams that you manage, educating [40:08] them on what's acceptable and what's not [40:10] because once they know what's acceptable [40:11] and what's not and they know what the [40:13] boundaries are, then they can create [40:15] with inside that, right? And I think [40:16] that's where things get really special [40:18] and that's where we see people's unique [40:20] and individual talents and creativity [40:22] and thought patterns really started to [40:24] emerge, right? Whether it's how we can [40:26] analyze data better, use this reporting, [40:28] pull abstracts from this to determine [40:30] how we're going to run, you know, ads or [40:32] campaigns or, you know, one of my [40:35] favorite things that I love to think [40:36] about is is letting again letting it [40:38] kind of analyze a system for red flags, [40:40] errors, antiquated data. Um, but again, [40:44] you can't do that until you set clear [40:46] boundaries and educate your team on [40:48] these things so that they can move [40:50] forward in a way that works best for the [40:52] company as a whole. Yeah. [40:54] Thank you. Okay, so we have a little [40:56] under five minutes left. I think we have [40:59] time for a couple questions, but before [41:01] we go to questions, quickfire, [41:04] name your favorite AI trend or tool. It [41:07] can be marketing related or not. Just [41:09] quick down the line. [41:10] >> Uh that fake band that came out that [41:12] like everybody was like, what is like it [41:15] was like a fake band that nobody could [41:17] find it on the internet. They were [41:19] racking up like like tons of plays on [41:21] like different streaming platforms and I [41:24] just thought that as a social experiment [41:25] was hysterical. [41:28] >> Uh mine is a bit futuristic. It is [41:31] humanoids boxing against each other in [41:33] garages uh in San Francisco. [41:38] >> I'm much more practical. I just I just [41:40] said my mine is fixer for email. I I [41:43] love it. [41:44] >> Yeah. I will say what I mentioned [41:46] earlier, Julius AI, just because I'm a [41:48] creative marketer, not a technical like [41:52] analyst. So when I saw what I could do [41:53] in minutes, I'm like, "Oh my gosh, like [41:55] this just saved me like hours in my [41:58] analyst hour." I was like, "I'm in [41:59] heaven." So yeah. [42:00] >> What was it? [42:01] >> Julius.ai. I don't represent it. I just [42:04] learned it last week. I've only used it [42:06] three times, but I'm already mind-b [42:07] blown. [42:08] >> Awesome. Thank you. And I saw a couple I [42:11] saw a couple questions up here. I don't [42:13] know if we have passed mics. So maybe [42:15] just stand up and shout or grab a mic. [42:20] >> Okay. [42:21] >> Uh [42:22] >> hold on. Let's [42:24] >> you mentioned NADN. What are some of the [42:27] most beneficial agenic workflows that [42:29] you've built using that platform? [42:32] >> Uh we I mean we use them a lot. Um so um [42:37] NN is a workflow um platform. Um so I [42:41] mean we're using it like everything [42:42] personally. So what we've challenged [42:44] each of our employees is any personal [42:47] workflows that you have, find a way to [42:50] um build a workflow to make it like [42:52] efficiently. So, um, so things that I [42:56] use it for, if I need to pull like info [42:59] from a spreadsheet and I need to take it [43:01] from a spreadsheet and maybe put it into [43:04] a CRM, maybe I need to put take it from [43:08] a spreadsheet, put it into a CRM, and [43:10] then put it into an email. Um, I'll I'll [43:14] run that through a workflow. Um, and put [43:16] that in Nad. Um, it maybe it needs to [43:20] come from like a department and then go [43:24] into a spreadsheet and into somewhere [43:26] else. Um, so I mean basic things like [43:30] that up to much more complicated things. [43:35] um our I mean we're closing real estate [43:37] transactions. So we use it um in our [43:40] transax teams um to like help with that [43:44] workflow through things in areas like [43:46] that where we may have gaps through [43:48] different systems that we use which is a [43:51] much more sort of complicated workflow [43:52] that we would use it um through there. [43:55] Um, so information flow, anytime there's [43:57] a flow of information between different [43:59] systems, we'll use it to bring that flow [44:02] from one system to another so that we're [44:05] not manually having to move over data.