Five years ago, enterprise technology roadmaps stretched across half a decade. Then they shifted to three years. Today, many organizations are planning just 18 months ahead, and even that horizon is getting shorter. As AI accelerates innovation, the challenge is no longer access to technology. The challenge is applying it effectively in a world that is changing faster than most organizations can plan for.

Perficient is not waiting for the dust to settle.

The global consultancy, with more than 7,000 advisors, engineers, and designers, has deep expertise in heavily regulated sectors such as banking, financial services, insurance, healthcare and life sciences, as well as auto and industrials. Working as a strategic partner to Fortune 100 companies, Perficient helps organizations become more data-driven, make better decisions, and move faster without losing their footing.

Its approach is grounded in a particular kind of pragmatism.

“Our focus is really on helping our clients understand the world around them and be very pragmatic in their approach to leveraging technology,” said Eric Walk, Vice President, AI Data Platforms at Perficient, “cutting through the noise and the buzzword bingo.”

That means treating AI as a practical tool rather than a destination, applying it to specific problems with clear outcomes in mind, and working iteratively with clients so they can adapt as conditions change.

“That type of agile iteration where we’re moving in weeks, not years, is the key to effective adaptation to the way the world is changing around us,” Walk said.

Building intelligence that can keep up

Speed alone does not improve outcomes. Organizations also need confidence in the direction they are moving.

Perficient uses a balanced intelligence model that brings together proprietary knowledge, open-source intelligence, and premium analyst research to create a complete view of the market. AI helps teams analyze large volumes of information and surface insights more quickly, and the firm is building internal AI agents to help consultants work more efficiently and deliver higher-quality work for clients. But the quality of those outputs depends entirely on the quality of what goes in.

“You can open up the world and have AI crawl the internet and look at any source of information, but you’re going to get results that reflect the internet,” Walk said. “It’s critical for us to ensure we have trusted inputs to produce trusted outputs.”

The role of trusted insight

That focus on trusted inputs is where analyst research earns its place in the model.

IDC’s research helps Perficient teams validate assumptions, challenge internal perspectives, and understand how markets are evolving in real time. It fills gaps in areas where consultants may not have deep specialization and gives them the external grounding to give clients advice they can stand behind.

“The more curated, more insightful, more thoughtful content that IDC provides is critical in helping us give our clients more effective advice and understand the world that we and they are living in day to day,” Walk said.

As AI becomes more integrated into consulting workflows, how teams access that insight is evolving too. Walk was among the beta testers for IDC Quanta, IDC’s new AI platform built on its proprietary research. For a firm already invested in making consulting more natively AI-enabled, the ability to query trusted IDC intelligence through a conversational interface, with every response grounded in specific research, points directly at the problem Perficient is trying to solve.

Turning disruption into advantage

Perficient is applying these same principles to itself, actively building AI-driven capabilities to improve how its own teams operate across research, delivery, and execution. Whatever it recommends to clients, it is already doing internally.

“We want to be out front and thinking about new ways of doing business and putting them into practice to drive outcomes for our clients,” Walk said. “At the end of the day, what matters is increasing revenue, reducing cost, and driving efficiencies in the right way to make both us and our clients successful.”

In a market that is not slowing down, that combination of trusted intelligence, honest advice, and a genuine willingness to lean into change is what separates the firms helping clients navigate disruption from the ones still trying to plan their way around it.

“We want to be the disruptor,” said Walk. “We want to lean into disruption.”

Christina Cardoza - Content Marketing Manager - IDC

Christina Cardoza is a Content Marketing Manager at IDC, where she specializes in brand content and social media strategy. With a background in journalism and editorial leadership, she has a proven ability to transform complex technology topics into clear, actionable insights.

2025年,智算云(AI IaaS)市场进入新一轮重构周期。行业演进的核心驱动力,已从单纯的模型训练需求,转向AI应用场景的快速扩张、底层技术瓶颈的持续突破,以及推理时代对算力组织方式的重新定义。训练、推理等工作负载与算力、网络、电力等基础设施开始深度耦合,推动AI IaaS从”GPU资源供给”逐步进化为涵盖算力底座、调度平台、模型服务与效能优化的全栈能力平台。

IDC认为,AI IaaS市场的竞争逻辑已发生根本性跃迁—— 从”资源供给”的单一维度比拼,转向以”算力底座、平台调度、模型服务与国产化生态”为核心的全栈能力竞争。这一转变,将决定未来三年市场格局的最终走向。

IDC最新发布的《中国智算云服务市场(2025下半年)跟踪》数据显示,2025下半年AI IaaS市场同比增速高达132.1%,市场规模达288.0亿元,2025全年市场规模达486.7亿元。(IDC定义下的AI IaaS覆盖GPU实例、AI专属集群以及相关AI存储服务,不含AI PaaS/SaaS层收入。)

算力资源成为AI产业核心壁垒

AI产业的竞争维度正在发生变化。过去行业更关注模型能力与产品体验,而当前,决定企业竞争上限的关键变量,正转向大规模算力资源的获取能力。对于头部基模厂商而言,pre-train、post-train、fine-tune、inference以及Agent工作流的持续扩展,都意味着对稳定智算资源的长期占用。高估值AI企业若想维持模型迭代速度与商业化增长,必须持续获得大规模、高稳定性的训练与推理资源支持。这一趋势也意味着,AI行业的竞争开始向更底层的物理资源延伸,包括土地、电力、数据中心以及先进芯片供应链。能够掌控这些资源的企业,将更容易建立长期竞争优势。除超大规模云服务商外,AI公司通常难以独立建设同等级基础设施,因此通过长期合作、预定集群乃至联合建设智算中心,正在成为行业主流路径。今年5月,Anthropic与xAI围绕算力资源展开合作,也进一步体现出“算力资源战略化”已经成为全球AI产业的重要趋势。

算力涨价潮背后的供需结构性矛盾

自2026年以来,AI算力产品价格持续上涨,反映出供需两端的多重刚性约束。需求端,AI模型规模和应用场景持续扩张,推理和训练算力需求同步激增。供给端,HBM高带宽存储、先进封装产能、电力基础设施等成为短板,算力边际增量有限,导致全球范围内算力涨价。国内市场上,今年3月,阿里云、腾讯云等头部厂商多轮上调AI算力及相关产品价格,反映出供需失衡的现实。涨价不仅体现在GPU集群租赁,还波及API调用、存储、网络等配套服务。IDC预计,短期内算力涨价压力仍将存在,推动企业优化算力采购策略,云服务商则需提升资源调度和能效管理能力以缓解供需压力。

推理时代到来,AI基础设施形态正在重塑

AI工作负载正在从“训练主导”逐渐转向“推理主导”。随着智能助手、AI Agent和自动化工作流逐步进入企业日常运营,AI调用已不再是单次问答,而是覆盖规划、检索、工具调用、代码执行、验证反馈等完整链路。一次任务往往对应数十甚至上百次模型推理请求,推理负载开始呈现高并发、低延迟、持续在线的新特征。

相比训练场景,推理更加关注:

  • 低延迟响应能力
  • 弹性扩缩容能力
  • 推理成本优化
  • 多模型混部调度能力
  • 边缘与中心协同部署能力

IDC数据显示, 2025年推理算力支出将首次接近训练算力支出。IDC认为,推理正在成为AI IaaS市场最重要的新增需求来源。

市场格局演化:“双阵营”结构形成

经过近一年的市场调整,中国AI IaaS市场逐渐形成“云服务商+运营商”双阵营竞争格局。云厂商延续了传统云计算时代在资源调度、产品体系和生态能力上的优势,并进一步向模型、推理服务与Token计费等AI链路延伸。

其中,阿里云依托灵骏智能计算集群,在大规模集群调度、异构算力支持、网络架构以及推理优化方面持续投入;百度智能云通过“芯云模体”体系,强化训练与推理一体化能力,在自动驾驶、金融、智能制造等行业推进落地;火山引擎则更多依托内部大规模推理场景积累,在推理成本优化与资源利用率方面建立差异化优势。运营商阵营则主要依托政企市场基础与国家战略资源,强化“算网融合”能力。其中,中国电信通过“息壤”平台推进全国算力调度;中国移动则通过“算力新动能行动计划”等持续强化算网协同能力。整体来看,市场集中度仍在提升,头部厂商在资金、资源获取和基础设施运营上的优势进一步扩大。

行业应用驱动市场结构重塑

当前,互联网与基模公司仍是GPU集群的主要采购方,但行业需求结构正在发生变化。互联网行业的音视频、电商、在线教育、社区平台等场景,正在持续扩大内容生成与推荐模型的推理需求;而基模公司则仍以训练和模型服务为主要资源消耗方向。与此同时,汽车自动驾驶场景和金融行业风控与智能投顾等场景也因其持续且庞大的计算需求,成为算力消费的重要力量。

国产化生态深化,供应链安全与自主可控成行业共识

外部供应链约束正在加速国产算力生态建设。过去,国产AI芯片主要受限于训练性能、生态兼容性以及集群稳定性,因此更多应用于边缘或特定行业场景。但随着推理需求快速增长,市场对“极致训练性能”的依赖有所降低,国产芯片在推理场景中的可用性明显提升。头部互联网公司、自动驾驶厂商等行业已在文本生成、图像识别、自驾仿真等多个场景中采用华为昇腾、百度昆仑芯、平头哥真武等国产AI芯片。供应链安全与自主可控,已经从政策导向逐步演变为行业共识。

IDC建议:AI IaaS竞争将进入“全栈能力”阶段

未来AI IaaS市场的竞争重点,将从单纯算力供给转向全栈能力竞争。

对于云服务商而言,需要持续强化异构算力布局、推理基础设施建设以及国产生态适配能力,同时围绕重点行业推出场景化解决方案,提升客户长期粘性。未来,API调用、Token计费、推理优化服务等新模式,预计将成为重要增长方向。

对于企业用户而言,则需要更加关注算力供给稳定性、推理性能、资源弹性以及整体成本效率。在推理需求快速增长背景下,按需租赁、弹性扩容与混合部署模式,有望在部分场景下逐步替代传统重资产采购模式,成为AI基础设施建设的新常态。

IDC 相关研究

围绕 AI 时代智算云与AI IaaS相关研究内容、技术能力与市场格局,IDC 将持续开展系统性研究,包括但不限于:

  • 《IDC Market Presentation:AI云存储市场进展及主流产品竞争力分析,2026》(即将发布)
  • 《IDC Market Presentation:中国企业智算云服务采购偏好调研,2026》(即将发布)

如需进一步了解 AI 时代智算云与AI IaaS相关研究内容,或咨询 IDC 在云计算、AI 基础设施及数字化转型领域的其他研究成果,欢迎与我们保持联系。

Qijin Chen

Qijin Chen - Senior Research Analyst

Chen Qijin currently works as a Senior Research Analyst for IDC China Enterprise Research. He specializes in analyzing and researching IT services, cloud professional/managed services, and DEOT services, providing industry analysis, forecasts, and market insights in related areas. In addition,…

2026年5月20日,全球大模型市场迎来“超级发布日”:谷歌在I/O大会连发Gemini 3.5 Flash、Gemini Omni两大新模型,阿里同步推出Qwen3.7-Max/Plus系列。中美巨头同周期密集迭代,直指Agent化、全模态、普惠高效三大核心方向,标志着大模型竞争从“参数竞赛”全面转向“能力落地+生态博弈”,全球大模型市场格局加速重构。

核心观点:技术路线齐向Agent,能力侧重各不同

谷歌主打全模态+极速性价比,锚定消费者Agent生态:谷歌推出多款 Gemini 系列模型, Gemini 3.5 Flash 速度更快、成本更低,综合能力升级,已是多款主流应用默认模型且用户体量庞大。同时谷歌依托专属平台与私人智能体布局核心战略,推动 AI 转变为全天候自主助手,深耕多类实用应用场景。全能模型 Gemini Omni 可融合多类媒介内容,具备多模态实力,还支持对话式视频编辑。谷歌未来战略以Antigravity 2.0平台+Gemini Spark私人智能体为核心,推动AI从 “工具” 进化为7×24小时自主执行的个人助手,全面渗透搜索、电商、创意场景。

阿里短短几月模型迭代至Qwen3.7,聚焦超长程Agent+部分开源:阿里 Qwen3.7 系列实力出众,Qwen3.7-Max面向智能体设计,在编程、推理等方面进行了能力更新,具备任务目标对齐能力,拥有 35 小时超长程自主执行能力,全新升级的Qwen3.7-Max即将通过阿里云百炼平台正式上线,全面对齐OpenAI与Anthropic API协议,并与Claude Code、OpenClaw及Qwen Code等主流智能体框架实现即插即用的无缝集成。视觉版 Qwen3.7-Plus 提升阿里视觉领域能力,打破海外多模态技术壁垒。此次发布是阿里云 “芯片-云-模型-推理” 全栈Agent化升级的一部分,该系列坚持部分开源的发展思路,着力搭建普惠型全球 AI 基础设施。

全球大模型迭代三大核心方向:可执行Agent、全模态融合、生态路线分化

从推理增强迈向 Agent 原生,行业竞争聚焦落地执行:  除了5月20日密集发布的模型,百度文心5.1也在本月稍早升级了智能体、知识、推理及深度搜索等方面的能力。可以看出Agent能力提升成为行业头部企业的共识,大模型发展下一阶段核心是智能体式思考,不再局限于基础逻辑思考。全球大模型企业将锚定智能体时代发展方向,优化模型长任务处理与工具调用能力,搭建平台助力智能体系统搭建,补齐长任务运行短板。整体行业趋势下,大模型正从单纯对话工具升级为自主生产力载体,评判标准也从精准度转为任务完成效率、运行稳定性等实战指标。

进阶全模态深度融合,角逐多模态技术高地:当前大模型赛道竞争正式进入全模态融合阶段,该能力已然成为旗舰模型必备配置。头部厂商打通文本、图像、音视频、3D 等多元内容交互链路,拓展创意生成新场景,加强视觉识别、逻辑推理与代码等能力,行业竞争重心也从单一能力比拼,转变为全模态协同能力与实际应用场景的全面比拼。

行业路线走向分化,普惠高效成为发展主流:大模型产业商业模式逐步分化,形成闭源商用与开源普惠或两者兼有的发展路径。谷歌依托闭源路线打造高性价比模型,凭借高速低成本优势抢占大众与企业市场,依托庞大用户体量构筑商业生态;阿里新模型坚守部分开源路线,降低开发使用门槛,汇聚全球开发者共建开放 AI 生态。两大阵营殊途同归,均聚焦能效比提升,摒弃盲目堆砌参数,以低成本轻量化部署作为核心发展优势。

四大路径助力产业破局:产品化、开源、生态和企业级特性

面对近期基础大模型带来的技术变革与市场机遇,IDC 给出如下建议助力中国大模型产业破局:

产品组合的广度与复杂性平衡:基础大模型产品组合的广度能带来优势,全球市场的头部厂商构建了高端推理,大规模部署, API 密集型长上下文场景,Agent场景,代码/软件工程场景,企业知识库/应用集成场景,前沿模型等专有模型。中国市场技术供应商可以考虑针对不同场景模型建立深度优化能力,而不是在所有场景中采取同质化的竞争策略;产品组合的广度也可能增加复杂性,可能增加客户在选择、集成、治理和总拥有成本(TCO)方面的复杂性。

开源与商业化策略:在大模型行业提速发展背景下,国内厂商持续优化开源与商业化布局。依托开源优势契合合规行业需求,兼顾成本、性能与市场需求,灵活平衡技术开放与商业落地,探索适配本土市场的可持续发展经营模式。开源模型采用开放且透明的设计,因此对那些要求模型具备高度透明度和合规性的受监管行业极具吸引力。对于将可解释性、可审计性和模型溯源列为采购要求的行业而言,能够检查模型权重、审核训练数据文档,并在无需依赖供应商的情况下修改模型行为,这本身就是一种竞争优势;另外,用户也在模型质量/推理深度、性能与预算之间寻求平衡, 中国技术供应商在以低成本触达用户的同时,还必须平衡模型质量/推理深度、速度与商业成果。

生态为王:中国大模型厂商需要与自有产品生态深度集成,并和全球广泛的上下游合作伙伴、系统集成商网络建立合作,并对全球范围内的开放协议积极加持;市场需广泛覆盖个人用户、开发者及企业级用户。

企业级特性纵深发展:在垂直领域,仍需配套企业级软件组合,通过生态合作伙伴网络加速行业知识的积累以及定制化解决方案的交付周期, 并配合专业服务。

IDC中国研究经理程荫表示,当前全球大模型正式迈入 Agent 迭代高速发展阶段,全球和中国市场技术供应商相继推出全新系列模型,开启行业新一轮竞争, IDC建议中国技术供应商从差异化产品、开源和商业化策略布局、全域生态搭建,到深耕企业级服务,多维度统筹发展,推动行业稳健提质、落地提速。

IDC在大模型与生成式AI领域有深入丰富的研究,欢迎广大技术供应商、行业用户、投资机构垂询。

请点击此处与我们联系。

Anne Cheng

Anne Cheng - Research Manager

Anne Cheng is a research manager in IDC China whose research focuses on the AI and big data markets. She collaborates with IDC's regional and global consulting teams and is involved in the business development of related markets. Prior to…

AI正在深刻重塑数据架构,并显著提升企业对其的关注度。随着生成式AI与Agent技术的爆发式发展,商业智能与分析、数据目录与血缘、数据质量评分、湖仓一体、数据治理等议题,已经成为超过40%组织的首要建设重点。

这一变化的深层逻辑在于,Data与AI正在形成一种前所未有的紧密交互关系,而非传统的运维流程管理。Scaling Law依然成立——高质量的数据支撑着上层AI与Agent的开发,而Agent本身又以数据库、数据湖、数仓、数据中台、数据分析平台为基础设施。更重要的是,Agent所产生的频繁交互,正在催生大量运行时数据和记忆的出现,这对数据的存储与管理提出了全新要求。

一、Data Agent:定义与趋势

IDC对Data Agent的定义是:利用Agent管理和治理数据,通过对话式或低代码入口实现精准查询、分析、决策,降低获取洞见的门槛。需要强调的是,Data Agent并非指向Agent工具本身,而是在广泛的数据场景中嵌入Agent能力,以实现更快速的数据集成、管理、开发、查询、分析和可视化内容生成。其核心覆盖场景包括数据集成、数据治理、数据发现、指标开发与知识管理、查询分析、Memory与上下文管理等。

从市场趋势来看,Data Agent已成为数据与分析行业未来的重要发展方向。IDC预测,到2028年,60%的中国500强企业将部署企业级Data Agent;到2026年,50%将部署数据分析Agent,以自动化日常任务、加速战略决策。

二、为什么Data Agent会成为趋势

要理解Data Agent的兴起,需要从政策、技术与市场需求三个维度来看。

首先是政策环境向好。2026年作为“十五五”开局之年,政府及央国企有望加大数字化预算。国家明确要激活数据要素潜能,深化“人工智能+”,这为Data Agent提供了坚实的政策土壤。

其次是基础模型的持续演进。过去大模型能力集中在文娱对话,而当前技术突破重点已转向数学、代码、长上下文理解与任务执行。这些能力对于数据开发和管理至关重要,使得Data+AI场景更加成熟可落地。

第三,Agent标准协议和框架的成熟正在加速发展。MCP、A2A等协议的快速普及,大幅降低了数据软件之间的互操作性难度,使得Data Agent的技术成熟度足以支撑大规模应用部署。

最后,企业技术架构正在螺旋式演进。2025年头部企业将预算优先投入湖仓一体与数据治理;随着基础设施完善,2026年企业希望通过Agent直接获取洞见、提升决策效率并降低人力成本。

三、为什么一定是Data Agent有更大需求

Agent可以在各个场景落地,但Data Agent会成为最先落地的方向,这可以从三个维度来回答。

从需求端看,企业的核心资产是数据。企业在对外合作和对内管理中沉淀了庞大的经营、财务、代码等数据资产。企业真正需要的,不是通用Agent,而是能够处理内部全流程的Data Agent。

从商业闭环看,Data Agent的盈利模式最容易实现。相比C端Agent仍处探索阶段,Data Agent面向企业客户,部署后可以快速显现效率提升。未来付费模式可能转向Tokens消耗或RaaS(结果即服务),且Data Agent的“每Token价值”比文字、视频内容更加清晰可衡量。

从技术迭代看,数据市场已发展数十年。所有企业都已接受并落地了各种数据底座。在已有IT基础之上,企业进一步“+AI”更容易实现,且能够保持技术延续性。

四、Data Agent在2026年会发展到什么程度

当前,Data Agent市场呈现出明显的供需错配:技术厂商的投入热情超过了客户的实际需求意愿。2025年厂商密集推出Data AI Agent产品,但在需求侧,企业仍处于基础设施建设和数据治理过程中,对新技术缺乏完整认知。这意味着2026年仍需厂商持续的市场培育。

OpenClaw的爆火远超预期,这一现象值得数据厂商深思。它反映出企业和用户真正需要的是“互操作性”和“主动性”。映射到数据层,这恰恰需要Data Agent来更好地接入和管理企业数据与Memory。

但必须承认,Data Agent的商业模式仍然未定。当前企业明确投入算力与数据,但对上层Agent功能是否单独付费尚未形成共识。Agent是附加价值还是下一代必需品?是单独付费还是默认“+AI”即为未来产品形式?市场倾向于后者,但尚未有定论。

此外,Data Agent最快落地的场景不是营销,而是商业分析。企业当前对AI的认可仍集中在“降本增效”上。Data Agent擅长执行内部数据工作,但复杂多变的营销场景仍需能力提升。

五、可能的抑制因素有哪些

在讨论抑制因素时,需要先明确一个判断:数据的高价值不会改变,但未来的话语权不一定仍在数据厂商手中。在AI逐步替代多种场景功能的趋势下,掌握数据的可能是AI软件而非传统数据产品。当前来看,做“中间层”(衔接数据与AI出口的综合引擎)的企业更有机会。

另一个重要判断是:Data Agent市场似乎不会出现“DeepSeek/Manus时刻”。数据市场更符合稳定增长路线,难以复现AI领域的轰动效应。数据厂商真正的护城河,在于多模态数据管理、治理与垂直行业经验。

此外,轨道偏移也可能带来不确定性。智慧办公等Personal Data产品可能颠覆传统SaaS交付逻辑,企业核心资产将从经营数据扩展为“经营数据+员工办公数据”,但市场目前缺乏对应产品。

最后,合成数据可能影响高质量数据的价值。虽然尚未完全成熟,但当合成数据的价值高于真实数据时(拐点取决于成本与准确率对比),数据资产价值可能被重构。

六、技术厂商应该怎么应对

面对上述趋势与挑战,技术厂商需要从四个方面着手应对。

第一,打造轻量化的数据平台。企业不会再经过漫长周期建设颠覆性数据底座,而是通过AI实现快速数据集成与应用,包括流式集成、Data Flow管理、本体与指标分析等。

第二,从被动工具型转向主动任务型。借鉴OpenClaw,让企业用户通过对话指令跨SaaS执行数据管理和分析。

第三,强化运行时安全。2026年预计会有更多厂商涌入Agent Infra赛道,保障Agent运行时安全与企业数据安全。

第四,关注Agent带来的数据变化。用户与Agent交互产生大量指令、记忆、中间态数据,厂商需尽快打造Data与AI之间的中间层,这将成为用户核心关注点。

七、市场应该如何选型

需要明确,Data Agent不是某一具体功能的代称,而是实现数据全流程AI自动化的总集。IDC调研显示,企业最希望建设的Data Agent类型包括:用户合规监管Agent、自动分析Agent、知识搜索Agent、动态执行优化Agent、自动决策Agent、Text2SQL Agent等。

在IDC《Data Agent MarketGlance 2026Q1》中,整体市场被划分为:Data Agent基础设施、数据集成与治理、平台厂商与通用智能体、轻量化工具与插件、垂直行业智能体、开源项目、安全方向。

从市场格局来看,传统Data Infra厂商正利用AI搭建数据开发入口;SaaS厂商基于客户资源拓展Data Agent能力;AI初创公司以轻量化插件快速对接数据格式,模仿Manus路径吸引用户。

分析师观点

IDC中国高级分析师李浩然表示,Data Agent将在2026年迎来快速落地。技术厂商需明确区分Data Agent与OpenClaw/AI办公软件的竞合关系,将自身开发经验沉淀为Skills和Mem-kit,加快轻量化部署。

与IDC同行,抢占Data Agent的战略先机

IDC长期追踪全球与中国Data Agent市场,已发布《Data Agent市场图谱2026Q1》《Data Agent市场预测,2026》《金融和零售行业Data Agent最佳实践》等报告,即将发布《中国Data Agent厂商评估,2026》《Data Agent开发平台技术能力评估》等重磅报告,并可为企业提供定制化场景评估与选型服务。

请点击此处与我们联系。

Leo Li

Leo Li - Senior Market Analyst

Leo Li is a senior market analyst on artificial intelligence (AI) and big data for IDC China. He conducts research and analysis on AI and big data for the China and worldwide markets. He is also involved in regional and…

在全球AI竞争格局中,中国AI模型在技术能力上已与北美差距缩小,但在海外市场渗透与用户信任方面仍面临显著挑战。当前,AI模型正从“工具”演进为“基础设施”,其能力已成为衡量企业技术竞争力与未来增长空间的核心指标。与此同时,Agent的繁荣并未削弱模型的重要性,反而放大了模型能力的差距。未来,谁掌握更强的模型、数据与Agent能力,谁就有机会成为下一代智能经济时代的核心平台。对中国厂商而言,开源、低成本部署与本地化能力,或将成为跨越“信任鸿沟”、在全球Agent建设中占据一席之地的关键路径。

本文核心洞察:

  • AI模型能力差距正在被放大,而非缩小
  • 中国模型已跨越“技术鸿沟”,但未跨越“信任鸿沟”
  • 海外企业拒绝中国模型的主因是长期支持、合规与成本,而非性能
  • 开源市场已超闭源,且“拿来即用”成为主流
  • 中国厂商的最大机会是成为全球Agent建设的“默认系统”

Agent繁荣之下,模型重要性不降反升

2026年以来,OpenClaw、Harness及Hermes等开源工具与智能体的出现,降低了技术供应商打造AI产品的门槛,推动了更多x-claw类产品的落地。尽管有观点认为“应用层会吞噬模型层”,但实际上,随着AI系统承担复杂任务与自主执行,模型能力的差距反而被放大。大模型正从“工具”演进为“基础设施”,AI模型能力已成为衡量企业技术竞争力、生态影响力与增长空间的重要指标。

模型迭代也直接推动了Agent的繁荣。2026年,全球近80%的企业已将智能体用于实际生产活动。中国当前平均部署12.84个智能体,计划年底达30.87个,但仍低于全球平均水平(当前23.5个,计划年底43.15个)。

全球市场的光与影:中国模型与北美仍有较大差距

从全球企业AI模型选择看,OpenAI(43.4%频繁使用,30.3%部分应用)占据主流,Gemini、Claude、DeepSeek、Grok、Doubao等亦被广泛采用。Doubao、GLM、Qwen、Hunyuan、Kimi在不同地区拥有客户群体。OpenClaw的出现及DeepSeek、智谱等厂商的快速迭代,也推动了中国模型的实际应用。

关键发现:剔除中国市场后,除DeepSeek外,其他中国模型在亚太其他地区、北美、西欧、中东/土耳其和非洲的广泛使用比例均不足10%。仅有DeepSeek超过15%,其余均低于10%。中国AI模型厂商在海外用户覆盖上面临较大挑战。

海外用户的顾虑:信任问题重于能力问题

对比用户选择因素,选择中国AI模型的企业更看重成本效益、透明度要求及多模态用例支持;选择北美模型的企业则更看重安全/合规、回答质量与性能效果。中国模型正在跨越“技术鸿沟”,但仍未跨越“信任鸿沟”。

进一步问询不考虑使用AI模型的原因,海外企业主要担心:中国AI模型厂商的长期支持水平、不满足安全合规要求、需要过多定制与微调、与系统不兼容、使用成本过高。而模型性能并非主要担忧因素,说明用户认知中,中国厂商的AI模型能力与北美差距较小。

开源:中国企业重构全球AI权力的工具

IDC数据显示,50%-60%的企业会应用开源模型而非闭源/商业化模型。尽管闭源模型更新会带来波动,但开源市场占比已超过一半。2026年Q2,DeepSeek V4及Qwen、GLM的持续迭代,将进一步推动企业用户采用。企业选择开源模型的原因包括:成本优势、数据隐私保护、深度定制化、技术自主性高及社区支持强大。

IDC还有以下发现:

  • 小模型获得更多选择:2025年,41.81%的开源模型参数量在2B-15B之间,58.5%的受访者选择该量级。
  • AI业务场景覆盖度为30%-40%:其中阿里通义系列占比最高,为41.2%。
  • 模型“拿来即用”比例升高:45%的受访者不进行调优直接部署,79.3%在调优后不会二次开源。
  • 开源模型落地的三种形态:40.9%部署到企业内部业务,32.0%开发为C端App/网页应用,27.1%与企业内部SaaS/PaaS/服务结合对外B端售卖。

未来机遇:Agent时代的“Android化”将是中国AI厂商的新机会。除了模型性能的追赶,中国AI模型的最大机会是通过开源、低成本部署、本地化能力和应用工具化,成为全球企业Agent建设中的默认系统和选择,以此获得全球更多信任。

分析师观点

IDC中国高级分析师李浩然表示,AI和Agent在全球各区域均进入快速落地阶段。即使经济环境下行,AI、安全、数据分析仍是企业主要投入方向。中国厂商AI模型在全球仍处于小范围部署和POC考虑阶段,随着模型性能、开源及产品生态的丰富,未来可能在全球获得更多机会。

本文相关IDC报告

  • 《全球企业级大模型采用率调研,2026》
  • 《中国AI开源市场研究报告(2025)》

进一步交流

如需获取完整报告或进一步探讨中国AI模型的全球化路径,欢迎与IDC联系。我们将基于一手调研与深度分析,为您提供更具针对性的洞察与决策支持。

请点击此处与我们联系。

Leo Li

Leo Li - Senior Market Analyst

Leo Li is a senior market analyst on artificial intelligence (AI) and big data for IDC China. He conducts research and analysis on AI and big data for the China and worldwide markets. He is also involved in regional and…

Your pipeline is growing. Conversion isn’t. That gap has a specific cause, and it’s not your demand generation.

In IDC’s recent expert panel, Addressing the Pipeline Conversion Gap, analysts examined why deal velocity and win rates are stagnating even as MQL volume climbs. The problem isn’t at the top of the funnel. It’s in what happens next.

They arrive having already evaluated options, checked pricing signals, and narrowed the field. Traditional qualification processes, built for a different era, are removing them before sales ever gets involved.

Today’s buyers are qualifying themselves. Your process just isn’t built for that.

The result is a filter that works backwards: organizations collect leads from browsers with no intent to buy while systematically losing their most qualified, motivated prospects.

“The funnel isn’t filtering for quality, it’s filtering for patience. And your best buyers have none.” — Heather Hershey, Research Director, Worldwide Digital Commerce

Five practices that push your best buyers out

AI has shifted information power to the buyer. Many qualification practices built to capture leads are now the reason those leads leave.

1. Gated content in an AI-first discovery environment

When decision-makers can get summarized answers instantly through AI tools, requiring registration for basic information creates friction without payoff. The problem has moved beyond conversion optimization. If your most valuable technical or educational content is inaccessible to AI indexing and answer engines, you risk disappearing from the places where B2B research now begins. The gate doesn’t just slow buyers down. It removes you from their consideration entirely.

2. Multi-step qualification chains

Traditional sales flows required multiple touchpoints before a prospect ever saw the product. Many buyers today want the opposite: the ability to self-evaluate first and engage with experts only when it’s worth their time. IDC research finds that high-intent buyers increasingly treat a mandatory demo sequence as a preview of what the purchase and implementation experience will be like. What they see is friction. What they do is leave.

3. Opaque pricing models

“Contact sales for pricing” once created negotiating leverage. Increasingly, IDC sees buyers interpret it as a signal that costs are inconsistent, negotiable in nontransparent ways, or deliberately difficult to benchmark. When prospects can compare alternatives through AI searches in minutes, withholding pricing information often accelerates movement toward competitors with cleaner commercial models , not away from them.

4. Requiring human contact for basic information

In IDC’s research, one of the most consistent buyer frustrations is being required to schedule a call to answer questions about integrations, security standards, deployment requirements, or product compatibility. When those answers are hard to find, buyers draw a conclusion: if pre-sale support is this cumbersome, post-sale support probably will be too. Self-service documentation isn’t a support cost. It’s where sales cycles now begin or end.

5. Repetitive discovery and qualification theater

Many organizations ask buyers to repeat the same background, pain points, and requirements across multiple calls with different representatives. What sales teams call thorough qualification is what decision-makers experience as a sign that your internal process matters more to you than their time does. Every redundant exchange reinforces that judgment.

The conversion gap isn’t a marketing problem

Most of the revenue challenges being attributed to marketing performance are symptoms of an outdated pipeline model. The organizations outperforming in this environment aren’t generating more leads. They’re removing friction for the buyers who already know what they want, and who expect to move on their own terms.

Closing that gap means changing some operating assumptions:

  • Let prospects self-educate before you bring them into human-led sales motions.
  • Make pricing, technical specs, and implementation details easy to find — not guarded.
  • Treat human engagement as something buyers opt into, not a gate you control.
  • Build for speed and buyer control. That’s it.

As AI intermediaries reshape how buyers discover and evaluate solutions, intent signals become harder to track. Qualification systems built for click-based, form-fill journeys are less effective when buyers expect answers before they ever identify themselves to a vendor.

Your competitors are probably making the same mistakes. The organizations that redesign for how buyers actually want to research and buy will capture the high-intent segment, the one that converts fastest and churns least.

Want to identify where your revenue process is creating buyer friction? Request a consultation with an IDC analyst.

IDC - -

International Data Corporation (IDC) is the premier global market intelligence, data, and events provider for the information technology, telecommunications, and consumer technology markets. With more than 1,300 analysts worldwide, IDC offers global, regional, and local expertise on technology and industry opportunities and trends in over 110 countries. IDC’s analysis and insight help IT professionals, business executives, and the investment community make fact-based technology decisions and achieve their key business objectives.

Google officially launched the Fitbit Air alongside the broader rollout of the Google Health platform, and while the hardware is getting most of the attention, enterprise and vendor watchers should keep their eyes on the software layer. This launch is less a fitness tracker story than a platform consolidation story, with significant implications for the broader health and wellness technology market.

The hardware: Deliberate, not derivative

The Fitbit Air is Google’s first screenless fitness tracker, priced at $99.99 and positioned squarely against the premium screenless wearables segment dominated by Oura and Whoop. The device is notably smaller than previous Fitbits and packs a meaningful sensor suite (heart rate, SpO2, HRV, irregular rhythm notifications, and automatic exercise detection) into a form factor designed to disappear on the wrist.

The deliberate absence of a screen is a strategic positioning decision, not a cost-cut. As the wearables market fragments, a growing segment of users prefers passive tracking devices that can be worn alongside a traditional watch rather than compete with one. The Fitbit Air isn’t the first screenless tracker to market, but it’s the first with a real chance to reach mainstream users, offering essentially the same health insights as premium incumbents at a fraction of the cost.

The lack of built-in GPS and NFC are notable omissions. Neither is a dealbreaker for the mass-market consumer Google is targeting, but both are worth flagging for more demanding buyers.

The real story: Google Health as a Platform

The hardware is secondary to what Google is actually building. The Fitbit app has been formally rebranded as Google Health, consolidating years of fragmented efforts across Google Fit, Health Connect, and Fitbit into a single platform. This is the kind of ecosystem rationalization that creates durable platform moats, and it matters for several reasons.

Google Health Coach, now out of public preview, is the differentiated layer. Built on Gemini, it functions as a personalized AI advisor across fitness, sleep, and health metrics; proactively surfacing insights rather than waiting to be queried. The coach delivers morning and evening summaries, post-workout analysis, and adaptive weekly fitness plans that update based on readiness scores and user context. In the US, it can connect to medical records, creating a longitudinal health data layer that goes well beyond what consumer wearables have historically offered.

What makes the coach genuinely interesting is its ability to incorporate life context, not just biometric data. When testing an earlier version as a new father traveling across multiple time zones with a sleep-training baby, the coach didn’t nag me to hit my workout targets. It recognized I wasn’t sleeping well, understood the circumstances I had shared, and offered practical alternatives such as a short walk, or pushing strength training to the next day. That kind of contextual flexibility is exactly what has been missing from wearables for years, and it’s where Google’s AI investment starts to show real differentiation over hardware-first competitors.

The platform also supports third-party device integrations via Health Connect and Apple Health, and Google intends to extend the experience to non-Google devices over time. The signal is clear: Google is less interested in selling you a tracker than in becoming the platform that ties all your health data together.

Competitive positioning and market implications

Several dynamics are worth watching. First, the Whoop comparison is apt but incomplete. The Fitbit Air takes the in-depth health analysis approach Whoop pioneered and opens it to the mass market at a budget price point. Whoop has a defensible customer base in enterprise wellness and athletic training; Google’s play is the mass consumer market plus enterprise wellness programs, an area where Fitbit has years of execution history. Notably, both represent a much larger total addressable market for Google than the competition it most directly displaces.

Second, Apple Health remains the obvious benchmark for platform completeness, and Google Health is clearly designed to close that gap on Android while making inroads on iOS. Limitations exist today, but Google has signaled plans to narrow the gap between Android and iOS experiences in the coming months.

Third, the subscription economics deserve attention. The Fitbit Air includes three months of Google Health Premium, after which the service runs $9.99/month or $99.99/year. Google is also bundling Health Premium into its broader AI subscription tiers (Google AI Pro and Ultra), a classic platform bundling move that mirrors its approach with cloud storage and productivity tools. This positions Google Health as a stickiness driver for the broader Google ecosystem, not a standalone revenue line.

One genuine limitation worth flagging

One architectural gap hasn’t received enough attention: although Google Health Coach is built on Gemini, it does not yet share context with the broader Gemini assistant. Users cannot tell Gemini “I’m traveling this week, pause my workout reminders until I get home” and have Health Coach act on it. The two systems currently operate in separate silos. For a company positioning Google Health as the intelligent, context-aware layer of its ecosystem, closing this gap should be a near-term priority.

The bigger picture

Google’s healthcare ambitions have historically outpaced execution, but this launch feels different. The coherence of the platform vision is clearer than it has ever been: a single app, an open third-party ecosystem, medical records integration, and an AI coaching layer all pointed at the same user experience. The Fitbit Air is the entry-point hardware, priced to drive volume. The data and subscription layer is the actual business, and the early signs suggest Google is serious about building it.

Jitesh Ubrani

Jitesh Ubrani - Director, Consumer Devices Research

Jitesh Ubrani is a Director at IDC leading a team of analysts within the Worldwide Consumer Device Trackers group, covering wearables, augmented reality (AR), virtual reality (VR), tablets, phones, PCs, gaming, and smart home devices, with a focus on market…

Ask a technology leader whether they’d rather have a fast answer or a right one and most will say both. Then tell you why they can’t have both. The AI tools are too unreliable for high-stakes decisions. The reliable research takes too long to surface. Pick your problem.

This is the speed-credibility trade-off, and it’s real in the sense that most organizations experience it daily. But it isn’t a law of physics. It’s a consequence of how enterprise research infrastructure has been built, and built badly, in most cases. The trade-off isn’t inherent to AI research. It’s inherent to AI research done with the wrong tools on the wrong foundation.

Where the trade-off actually comes from

The speed problem and the credibility problem look like opposite sides of the same coin. They’re not. They have different causes and different fixes.

Speed failures come from delivery infrastructure. Research that lives in a portal, a PDF, or an inbox takes time to find, time to read, and time to apply. Even when the underlying research is excellent, the process of getting it to the decision slows everything down.

Credibility failures come from the research base. When an AI tool generates an answer from the public internet, it has no way to verify that the answer is current, proprietary, or traceable. Hallucinated citations and outdated data are symptoms of an ungrounded model, not of AI speed per se.

Organizations that treat these as a single problem end up solving neither. They slow down to get credibility, or speed up and lose it. The ones that separate the two problems can address each on its own terms — and in doing so, discover that the trade-off disappears.

Speed without credibility is the symptom, not the disease

The AI credibility crisis is real and well-documented. Public AI tools produce confident, fluent, incorrect answers with a regularity that has made enterprise technology teams legitimately cautious about using them for anything that matters. That caution is appropriate. It’s the conclusion some organizations draw from it that isn’t.

The conclusion: AI tools are fast but unreliable, therefore slowness is the price of defensibility. This framing treats ungrounded AI as the only kind of AI, and the public internet as the only available research base. Neither is true.

The credibility problem is not that AI produces answers quickly. It’s that AI produces answers from sources that can’t be verified. Fix the source, ground the model in proprietary, cited, traceable research, and the speed advantage of AI delivery becomes an asset rather than a liability.

“An AI backed by IDC’s research gives me a lot more confidence in the answers.” — Phillip Langeberg, CTO, The Resorts Companies

Confidence and speed aren’t in tension here. They’re both present, because the research base they draw from is trustworthy by design, not by process.

What trustworthy by design actually means in practice

There’s a meaningful difference between research that is credible because someone checked it and research that is credible because the system that produced it is built on verified, proprietary data.

The first model, credible by process, is what most organizations rely on today. An analyst reads the AI output, checks it against source material, and flags anything that looks wrong. This is better than nothing. It’s also slow, labor-intensive, and dependent on the analyst’s availability and judgment.

The second model, trustworthy by design, is what changes the trade-off. When every answer runs through a multi-agent verification layer checked against proprietary research, when sources are cited and linked, when the reasoning behind each output is visible and expandable, the checking doesn’t require a separate human step. It’s built into the architecture. The answer arrives fast and defensible, not fast and then verified.

“Where it used to take weeks to draw conclusions from hundreds of reports, I can now do that in minutes.” — Mark Terranova, Director, Worldwide Analyst Relations, Kyndryl

Weeks to minutes. With cited, traceable outputs. That’s not a description of accepting a speed-credibility trade-off. It’s a description of what happens when the trade-off is structurally eliminated.

The three conditions that eliminate the trade-off

The speed-credibility trade-off disappears when three conditions are met simultaneously. Miss any one of them and the trade-off returns.

The research base has to be verifiable at the source. Not scraped from the public internet and then filtered, but proprietary, current, and built on methodology that can survive scrutiny. Six decades of IDC research represents exactly that kind of foundation — the kind that gives AI outputs something defensible to stand on.

The delivery has to be embedded where decisions happen. Even the most rigorous research fails the speed test if it’s sitting in a portal. Intelligence that arrives in the workflow, in the collaboration tools, in the AI environment where the team is already working, closes the delivery gap without requiring anyone to go looking.

The outputs have to be traceable. Speed without traceability is just a faster path to the same credibility problem. Every answer needs a visible source, an expandable reasoning trail, and a citation that can be put in front of a CFO or a board without apology.

Meet all three conditions and the choice between fast and defensible stops being a choice.

Why this matters for how you build

The speed-credibility trade-off has become a planning assumption for many enterprise technology teams — a constraint that shapes budgets, workflows, and governance frameworks. Accepting it as fixed has real costs: slower decisions, underutilized research investments, and an organizational posture that treats caution and speed as permanently opposed.

The more productive assumption is that the trade-off is a design problem. And design problems have solutions.

IDC Quanta is built on all three conditions: six decades of proprietary IDC research as the foundation, AI delivery embedded in the tools where decisions happen, and multi-agent verification that makes every output traceable before it reaches you. 97% of early access customers rated it as meeting or exceeding expectations. The platform launches this summer.

If you want to see what eliminating the trade-off looks like in practice, the link below takes you there.

Ryan Smith - Content Marketing Director - IDC

Ryan Smith is the Director of Content Marketing at IDC, where he leads brand-level content and social media strategy, aligning research insights with compelling storytelling to engage technology decision-makers. With a background in both IT and marketing, Ryan brings a unique blend of technical understanding and creative strategy to his work. He’s also a seasoned storyteller, speaker, and podcast host who believes the right message, told the right way, can drive both trust and transformation.

When drone strikes hit data centers in the UAE and Bahrain, disrupting cloud regions, it was more than an infrastructure crisis. It was a stress test the industry wasn’t prepared to pass. The old playbook of cybersecurity, disaster recovery, and backup no longer covers the necessary ground. Business continuity now carries a new mandate. Organizations that grasp this shift early will be the ones still operating when the dust clears.

What the conflict revealed about resilience

The numbers from the region are stark. During the escalation period, UAE organizations reported more than 800,000 cyberattacks per day, a figure documented across regional incident tracking and government disclosures. More than 150 hacktivist and cyber incidents swept across the region. Oracle was named as a target by Iran’s IRGC; Microsoft, Google, and Nvidia received direct threats. Subsea cables and Gulf digital infrastructure faced escalating risk. Airspace closures disrupted not only aviation but also connectivity, technology systems, and daily operations.

At the same time, physical access to offices, data halls, and SOCs became impossible for many organizations. This was not because their systems were compromised, but because security restrictions, travel advisories, and employee displacement locked recovery teams out of the buildings where they needed to work.

This is the defining lesson of this conflict: Cyber and physical threats no longer take turns. They converge.

Five things war conditions change for business continuity

  1. Premises denial is now as critical as cyber compromise. Even when systems are intact, staff may be unable to reach offices, SOCs, plants, or branches due to security restrictions, curfews, or evacuation orders. Physical access failure is a continuity failure.
  2. Recovery teams become the bottleneck. Displaced personnel, degraded telecom access, and broken leadership chains make it harder to approve failover, validate clean backups, or execute safe recovery. People are the constraint now, not just technology.
  3. Physical and cyber events are the same event. A missile strike, cable cut, or power disruption opens exactly the window attackers exploit, ransomware, wipers, credential abuse, and disinformation running in parallel.
  4. “Availability” means more than uptime. In conflict conditions, the real continuity question is operational: can the organization still deliver critical services to customers, citizens, patients, or counterparties at the minimum level that keeps them functioning?
  5. Geographic redundancy can fail if it’s too correlated. If your primary and DR locations share the same regional risk envelope, you may technically have redundancy, but operationally you’re still exposed. Diversity of risk, not just diversity of location, is what matters.

The new framework: ResOps

Organizations navigating this moment with confidence are building what IDC defines as Resilience Operations (ResOps): a continuous, operational discipline that spans cyber, infrastructure, and business layers simultaneously.

ResOps is not a product or a checklist. It is the organizational capacity to maintain essential services when faced with degraded networks, dispersed staff, and inaccessible facilities, all at once, in real time.

The value proposition is direct: stay remotely operable, controlled, and clean under conditions that would break a traditional continuity model.

Four pillars for building a resilience offering in the region

For security vendors and MSSPs operating in the Middle East, IDC’s research identifies four areas where providers need to structure their capabilities:

  1. Protect: Harden and segment everything that can be targeted, digitally and physically. This includes identity and access resilience, data and workload protection, and physical and environmental security.
  2. Isolate: Assume compromise and loss of premises. Capabilities here include clean room recovery, cyber vaulting, out-of-band secure administration, forensic readiness, and recovery orchestration.
  3. Operate: Keep essential services running. This requires remote SOC operations, workforce continuity planning, alternative connectivity, crisis command centers, and defined minimum viable service thresholds.
  4. Prove: Test, learn, and demonstrate resilience through simulations, tabletops, recovery testing, sector runbooks, and third-party assurance.

Where spending is going

IDC’s Future Enterprise Resiliency & Spending Survey is clear: security and resilience budgets are among the most protected in the enterprise, regardless of broader economic conditions.

Cyber recovery and cyber resilience are expected to see significant investment increases across the META region in 2027, pushing META security spending to nearly $13 billion by end of 2026. Providers who position around resilience operations, not point security products, are the ones who will capture this spend.

The bottom line for providers

The Middle East war is not an isolated event. It’s a preview of what geopolitical risk looks like for enterprise technology in the years ahead.

Providers who win in this environment are those who help organizations answer one question with confidence: when everything goes wrong at once, can you still deliver?

Build your portfolio around that answer. Give your customers the evidence they need to act and the path to get there.

Shilpi Handa

Shilpi Handa - Associate Research Director (META), IDC

Shilpi Handa is an associate research director at IDC, with responsibility for the Middle East, Turkey, and Africa cybersecurity practice. Her core research coverage revolves around cybersecurity, with a focus on network security, cloud security, application security, and security operations.…

AI is redrawing the rules of the partner ecosystem faster than most organisations can adapt. Last week, Stuart Wilson, IDC’s Senior Research Director for Partnering Ecosystems, and Andreas Storz, Senior Research Manager in the same practice, shared IDC’s latest research on what that means in practice for vendors, partners, and distributors operating across Europe and beyond. Drawing on survey data from more than 1,000 established partners, direct feedback from IDC’s European Partner Advisory Board, and real-world vendor examples, they made the case that the ecosystem is not simply evolving: it is being structurally reset. Here is a brief overview. The full recording is available on demand. 

How AI is eroding traditional partner revenue streams 

Stuart opened with a finding that will resonate with anyone tracking partner economics right now: AI is systematically compressing the lifecycle phases where partners have historically earned the most. Implementation, integration, and basic support are not disappearing, but they are becoming thinner and, in a growing number of cases, absorbed directly into vendor platforms. IDC has documented specific examples of how this compression is already playing out at scale, with leading vendors publicly committing to timelines and automation levels that would have seemed ambitious just 18 months ago. 

What makes this moment different from previous platform shifts is the speed and simultaneity of the impact. AI is hitting vendor economics, partner margins, and customer expectations at the same time. Partners who are waiting for the dust to settle before repositioning are likely to find the window has already closed. 

Where partner value is growing: advisory, AI governance, and outcome-based services 

The compression of execution-heavy activities does not mean the overall ecosystem opportunity is shrinking. IDC’s data points clearly to a redistribution of spend toward higher-order roles: AI solution design and agent creation, governance and compliance services, industry advisory, data engineering, and reusable marketplace IP. These are areas that reward deep domain knowledge and customer trust rather than delivery capacity. 

Customers are also changing how they expect to be served. Rather than relying on a single partner to cover the full lifecycle, they increasingly want a coordinated network of specialists. That shift has direct implications for how vendors structure their ecosystems and how partners think about collaboration rather than competition. The recording covers the full breakdown of where IDC sees demand growing and shrinking, and what partners are doing today to get ahead of it. 

Agentic marketplaces and the new partner go-to-market playbook 

Andreas Storz walked through a structural shift in how technology solutions are discovered and bought. Marketplaces are moving inside products, and IDC is seeing real evidence that customers are making decisions before a formal procurement process ever begins. This compresses buying cycles, introduces new buyer personas including business users and domain specialists who are not traditional IT buyers, and moves partner influence upstream into phases where most partner programs have little presence today. 

Partner Advisory Board members were frank about what this looks like from the front line. One dimension that generated particular discussion was the growing scrutiny customers apply to every new AI investment: 

“Customers scrutinize every purchase order. We are having to prove the ROI to the last penny before new AI work is approved.” — IDC Partner Advisory Board member, November 2025

Co-sell models are adapting accordingly. The shift away from field-led selling toward digital and telemetry-driven motions is not a future state: it is already shaping how the most forward-leaning vendors are structuring partner engagement today. The recording covers what that looks like in practice and what partners need to do to remain visible in these new buying journeys. 

Why vendor partner programs need a fundamental redesign for the AI era 

The structural conclusion Stuart and Andreas reached is that most partner programs in operation today were designed for a world that is rapidly ceasing to exist. The incentive structures, metrics, and engagement models built around resale transactions and implementation milestones are misaligned with where ecosystem value is now being created. Vendors that do not address this gap will find themselves losing the partners best positioned to deliver AI-driven outcomes to customers. 

The research points to a dual imperative: accelerating existing partners toward AI-centric delivery models while simultaneously cultivating a new generation of AI-native partners who bring differentiated industry IP and a very different set of expectations around how vendor relationships should work. How to run both strategies in parallel, without letting either undermine the other, is one of the more complex programme design challenges IDC is helping clients navigate right now. 

“AI-native competitors without legacy delivery models are coming. If we don’t pivot, we will be disrupted.” — IDC Partner Advisory Board member, November 2025 

The Q and A that followed also surfaced sharp questions on how AI model providers are disrupting established alliance hierarchies for global systems integrators, and whether the net effect of all this change will be a more consolidated or more fragmented ecosystem. Stuart’s answer to the latter was more nuanced than a binary either/or, and worth hearing in full. 

The complete recording includes data from the IDC EMEA Partner Survey data (N=1,001), a detailed breakdown of the dual partner strategy framework, and a live Q and A with both analysts. If the topics covered resonate with your ecosystem strategy, IDC’s Partnering Ecosystems practice offers advisory support, custom research, roundtables, and strategic workshops tailored to vendors, distributors, and partners navigating this transition.  
 
Our experts are always happy to continue the conversation. Simply reach out via the contact form.  

Stuart Wilson

Stuart Wilson - VP, WW Partnering Ecosystems, Alliances and Channels

Stuart Wilson is VP, Worldwide Partnering Ecosystems, Alliances & Channels at IDC. He leads IDC's research and advisory agenda focused on partnering ecosystems, strategic alliances, channels, marketplaces, and ecosystem-led growth. He currently leads IDC's EMEA Partnering Ecosystems research program and…
Andreas Storz

Andreas Storz - Research Director, Software Channels and Ecosystems

Andreas Storz is Research Director for IDC's Software Channels and Ecosystems program. Based in the US, Andreas focuses on the evolution of go-to-market models, new digital value chains and the wider impact on partner ecosystems, exploring how current and future trends will…