Google 在其官方栏目 Ads Decoded 的最新一期(时长 25:11)中,由 Google Ads 产品联络人 Ginny Marvin 对话搜索业务数据科学与用户体验研究(UXR)副总裁 Shivani Mohan,讨论搜索正在经历的两条并行变化:一是搜索体验与产品能力本身的迭代,二是用户搜索习惯的迁移。节目明确点出,这两条变化正在直接影响广告生态的底层逻辑,因此值得广告主与优化师优先关注。
Mohan 及其团队的核心工作,是对数十亿量级的用户搜索行为进行观测与研究,并重点追踪 AI Overviews 与 AI Mode 推出之后用户行为发生的偏移。这类研究属于平台侧的"幕后"洞察:它不只是产品功能的迭代说明,更是理解流量结构、意图分布与交互路径如何重构的一手依据。对于依赖搜索流量做用户获取(UA)与增量提效的广告主,这类行为研究往往先于投放产品的正式更新,具备前置参考价值。
节目特别强调了长句式查询(long-form query)趋势对企业的意义。当用户从短关键词转向更接近自然语言、条件更复杂的表达时,单条查询背后承载的意图更丰富,但意图的颗粒度也更难用传统关键词匹配去覆盖。这直接牵动投放策略的调整方向:关键词结构需要重新分层,素材与落地页需要更强的语义承接能力,而转化路径与归因分析(attribution)也要能够识别"更长决策链"中的多点触达,而非只归功于最后一次点击。
从实践角度看,这期内容的落点在于提醒广告主把"搜索"重新理解为一种持续演化的用户行为,而不是一组静态的关键词库。相应地,预算分配、出价与素材测试节奏、以及衡量体系(包括对 ROAS 与 LTV 的口径校准)都应保留对搜索侧变化的响应空间。对于出海与效果类团队而言,尽早建立对长句式查询的监测与内容承接能力,是把握下一轮搜索流量红利的关键动作。
需要说明的是,本文为播客节目导语,主要披露了议题框架与研究方向,并未给出具体的百分比或金额类数据;更细的研究结论需收听完整节目(25:11)获取。
值得关注的是,这次访谈由负责数据科学与UXR的搜索团队出面,而非广告产品团队,这本身就传递出信号:查询形态先于广告形态变化,广告侧的适配是滞后的。长查询占比上升,意味着关键词匹配、否词体系以及落地页意图识别所依赖的短词假设正在松动,账户结构的历史经验值也需要重新校准。放在AI搜索入口争夺的背景下看,Google 主动公开用户行为研究,更像是在AI搜索商业化节奏明确之前先与行业对齐预期,而非单纯的科普。
对UA与变现团队而言,搜索流量的结构变化通常先反映在转化路径上,而非曝光量上;这也是该内容在当下比常规产品更新更值得一读的原因。
This TikTok for Business landing page promotes a limited-period new-advertiser incentive program offering up to USD 6,000 in matched ad credits. Three self-serve tiers are presented: 50% off first 30-day spend (up to $200), dollar-for-dollar match up to $500, and match up to $6,000. A separate headline offer advertises Spend $100/Get $100, Spend $500/Get $500, and Spend $1,500/Get $1,500 plus 1-to-1 expert onboarding support. Registration requires email or phone sign-up with 2-step verification. Eligibility is restricted to new SMB self-serve accounts, excluding agency, Business Center, and TikTok Shop accounts; credits post within 7 days of hitting the highest threshold or after the 30-day window.
Attribution fails because platforms self-report, data is siloed, identity breaks across surfaces, and traditional models answer the wrong question. The same purchase is often claimed by three or four platforms, inflating attributed revenue 2–4x and distorting ROAS and budget decisions. For ad ops leaders, the fix is not another dashboard but an independent, deduplicated measurement layer with cross-surface customer ID, full coverage of mobile, web, CTV, and retail media, plus incrementality testing. Clean attribution is also the prerequisite for reliable AI bidding and budget automation; broken inputs scale misallocation faster.
TikTok is offering new advertisers up to $6,000 in ad credits through a tiered spend incentive ($100/$500/$1500) that includes 1-to-1 expert support at the top tier. However, eligibility is restricted to new SMB self-serve accounts, and credits expire. Alongside the offer, TikTok has rolled out several ad tech innovations—Symphony AI creative suite, Streaming Ads, Agentic Hub, Market Scope, and new MMM data—that provide actionable opportunities for testing and scaling performance. Ad ops teams should review eligibility criteria carefully and consider leveraging these tools to maximize ROI during the promotional window.
TikTok for Business is promoting tiered first-30-day ad credit incentives for new self-serve SMB advertisers: spend $100 get $100, $500 get $500, or $1,500 get $1,500 plus 1-to-1 expert onboarding support, with a headline entry offer of 50% off first-month spend up to $200 and matching credits up to $6,000. Credits are applied within 7 days of hitting the highest threshold or after the 30-day accumulation window. Eligibility excludes agency, Business Center, and TikTok Shop accounts, and limits one ad account per new TikTok For Business account. For ad ops teams, these terms dictate budget pacing, account structure, and credit-expiry planning.
The open internet presents unique challenges for performance advertising: fragmented identity, closed first-price auctions, and non-stationary supply. Moloco's CARA compound architecture tackles this with six integrated technical domains—Campaign Automation, Supply, Ad Recommendations, Bidding, Creative, and Signals—running on a unified ML infrastructure. Key insights for ad ops: the system continuously learns from every interaction, uses knowledge distillation to serve real-time predictions under 10ms latency, and validates improvements through rigorous live experiments. In 2025, 65 validated model updates reduced CPA by 17% and improved ROAS by 27%. The key takeaway: compound AI architectures that connect prediction, bidding, creative, and data can unlock measurable performance gains beyond walled gardens.
TikTok's Lorenzo Bierhuys argues that equating performance marketing with bottom-of-funnel capture is outdated; performance should instead mean data-driven, measurable marketing applied across the entire journey. TikTok frames discovery as the start of an infinite purchase loop, citing that 61% of users purchase after seeing an ad, 83% after using TikTok for inspiration, and 41% after a 'where to buy' search. Case evidence backs the full-funnel case: Mejuri +11% ROAS with 88% more efficient CPA, Tatti Lashes 13x ROAS, Conforama 239% CTR lift, and Ooh & Aah Cookies 8.81 total ROAS. Takeaway for ad ops: plan around intent signals first, then select capture routes — TikTok Shop, Pixel, App Events API, or Lead Generation ads.
AppsFlyer now attributes AI-driven app traffic on both sides of the funnel: earned organic discovery across 8 AI answer engines (ChatGPT, Claude, Perplexity, Gemini, Grok, Copilot, Manus, DeepSeek) and 13 search engines, plus paid campaigns via a new ChatGPT Ads integration using OpenAI's Conversions API. For ad ops teams, the practical win is deduplication and incrementality: SEO and AEO performance can be viewed side by side, and paid versus organic splits on the same platform (e.g., ChatGPT) reveal the true lift of ad spend rather than cannibalized organic conversions. Setup is low-lift, leveraging existing OneLink, App Links/Universal Links, and Smart Script infrastructure, with no added cost.
Meta’s Advertising Week New York updates center on AI-driven campaign automation and faster creative production. Ads Creative Studio is expanding with performance insights. Meta AI business assistant gains creative insights, deeper business memory, and agentic abilities to plan, create, optimize, and schedule tasks, with advertiser permissions and approvals. AI-enabled audience discovery arrives in Detailed Targeting, and Customer Lifecycle Strategy helps prioritize new customers. Generative AI video expands for static assets, catalog product video, and smart recommendations. Direct checkout tests launch in Meta Business Agent on Messenger, Meta AI, and Muse. For ad ops, these tools promise efficiency, but governance and testing remain critical.
平台方将 AI Mode 广告扩展至澳大利亚、加拿大、印度、肯尼亚、马来西亚、新西兰、尼日利亚、巴基斯坦、菲律宾和新加坡,并计划于 Q4 2026 进一步覆盖巴西、墨西哥和英国;同时在美国上线面向追问场景的多轮对话广告。关键数据方面,AI ...
Google Ads 在最新一期《Ads Decoded》中系统介绍了新一轮出价与预算工具,主张以工具升级带动假日旺季投放策略的整体重构。节目由 Google Ads 产品联络人 Ginny Marvin 与广告产品总监 Tal Akaba...
Google 在美国和加拿大推出 Financial Services Verification,在身份验证基础上增加金融广告主资质核验,扩展现有金融诈骗防护计划。该计划覆盖42个国家、约20亿互联网用户,美加已有超97%的 Google ...
Google Ads 正式推出「渠道优先级」与「预算分配」两项可选控制功能,让广告主可以自主决定预算在 Google Search、Search partners、Google Display Network 等渠道间的流向与先后顺序。新功...
Google 在十月 Demand Gen 更新中强调,Demand Gen 的投放效果上限取决于数据质量,数据强度越高,测量准确度与 ROI 表现越好。基于内部归因分析,借助 Google Tag Gateway 强化数据基建的广告主,其...
在 Google Ads 的《Ads Decoded》最新一期中,Ads Product Liaison Ginny Marvin 与 Demand Gen 及 YouTube 效果广告产品经理 Alejandro Osio 对谈,指出 D...