Competitive Landscape Analysis: A Practical Method竞争格局分析方法：从市场范围、玩家细分和证据评估到趋势判断、情景推演与行动优先级

What Is Competitive Landscape Analysis?什么是竞争格局分析？

The process begins with a decision and ends with an owner, an action, and a review date. Its purpose is not to decorate a slide with logos. It makes assumptions visible: which customers and jobs define the market, which alternatives buyers seriously consider, why players belong in the same group, what supports each claim, and which changes would invalidate the conclusion.

流程从一个决策问题开始，以负责人、行动和复核日期结束。目的不是用品牌标志装饰幻灯片，而是让假设显性化：哪些客户和任务定义市场，买方真正考虑哪些替代方案，玩家为何属于同一组，每项主张由什么支持，以及哪些变化会使结论失效。

Work through the fields below to define the market, identify players, select comparison dimensions, and document evidence. If you first need definitions, market structures, and broader examples, read the competitive landscape guide. Use competitor analysis when selected firms require deeper examination. The verified player set and dimensions can then become prepared inputs for the InfiniSynapse comparison workflow.

可按照下方字段定义市场、识别参与者、选择比较维度并记录证据。如果需要先了解定义、市场结构和更广泛的案例，请阅读竞争格局指南；如果部分企业需要更深入的研究，可采用竞品分析。经过核验的参与者集合和比较维度还可作为InfiniSynapse 对比工作流的准备输入。

Step 1: Write a Decision-Led Analysis Brief步骤一：编写由决策驱动的分析简报

Before collecting data, write a one-page brief. State the decision, decision owner, audience, geography, customer, job to be done, time horizon, delivery date, resources, and confidence required. “Map the AI software market” is too broad. “Choose which evidence-management workflow to serve for European mid-market product teams during the next twelve months” creates a workable boundary and a reason for every later inclusion.

收集数据前先写一页简报。说明决策、负责人、受众、地区、客户、待完成任务、时间范围、交付日期、资源和所需置信度。“绘制 AI 软件市场”过于宽泛；“选择未来十二个月面向欧洲中型产品团队服务的证据管理工作流”则建立了可操作边界，也为之后每次纳入提供理由。

Add explicit exclusions. A regional study may exclude companies that cannot sell or support customers there. A product-entry study may exclude adjacent providers that do not solve the selected job. Record exclusions so reviewers can challenge them without silently changing scope.

还要写明排除规则。区域研究可以排除无法在当地销售或提供支持的公司；产品进入研究可以排除不能解决所选任务的相邻供应商。记录排除项，让审阅者能够提出异议，而不是悄悄改变范围。

Step 2: Build the Player Universe from Buyer Alternatives步骤二：从买方替代选择建立玩家全集

Start with the buyer’s switching set rather than a “top ten” list. Include direct competitors that solve the same job similarly, indirect competitors that solve it differently, adjacent entrants, service providers, internal workflows, spreadsheets, delay, and doing nothing when those are credible choices. This view reveals the real pressure on adoption and pricing.

从买方可切换选择集合开始，而不是从“十大公司”名单开始。纳入以类似方式解决相同任务的直接竞品、以不同方式解决任务的间接竞品、相邻进入者、服务提供商、内部流程、电子表格，以及在可信情况下的推迟或不采取行动。这能揭示采用和定价面临的真实压力。

Use multiple discovery routes: customer interviews and win-loss notes; procurement shortlists; partner catalogs; official product pages; regulatory registers; public filings; standards bodies; trade associations; and reputable industry databases. Search results can surface candidates, but they are not proof that a company belongs in the market. Confirm the entity, offer, geography, customer, and current status with dated evidence.

使用多条发现路径：客户访谈与赢单/输单记录、采购短名单、合作伙伴目录、官方产品页、监管登记、公示文件、标准机构、行业协会和可信数据库。搜索结果可以发现候选者，但不能证明某家公司属于该市场。应使用带日期的证据核实实体、产品、地区、客户和当前状态。

Maintain a candidate log with included, watch, and excluded statuses plus a reason. Avoid choosing only famous firms or companies that resemble your own product. The objective is coverage of buyer choice, not a flattering comparison set.

维护候选台账，设置纳入、观察和排除状态，并写明理由。不要只选择知名公司或与自身产品相似的企业。目标是覆盖买方选择，而不是制造有利于自己的比较集合。

Step 3: Segment Players with Observable Rules步骤三：用可观察规则划分玩家

A segment is useful only when membership predicts something relevant: buyer, economics, route to market, product architecture, delivery model, regulatory burden, or competitive response. Create a first-pass classification, assign every included player, and review ambiguous cases. If half the market belongs in “other,” the scheme is not explaining the market.

只有当成员归属能够预测相关事项时，细分才有用，例如买方、经济模式、进入市场路径、产品架构、交付模式、监管负担或竞争反应。建立第一版分类，为每个纳入玩家分组，并复核模糊案例。如果一半市场都被放入“其他”，说明分类没有解释市场。

Do not copy vendor labels. Write a membership rule that two analysts could apply independently. When disagreements occur, improve the rule or mark the player as spanning segments.

不要复制供应商标签。应写出两名分析人员可以独立应用的成员规则。发生分歧时，要么改进规则，要么标记该玩家横跨多个细分。

Step 4: Select Dimensions That Explain Buyer Choice步骤四：选择能够解释买方选择的维度

Choose dimensions after defining the decision and segments. Useful dimensions may include target customer, supported job, product breadth, capability depth, implementation effort, outcome evidence, price architecture, distribution reach, ecosystem integration, service intensity, trust requirements, and switching cost. Each dimension needs an operational definition and permitted evidence sources.

应在定义决策与细分后选择维度。有效维度可以包括目标客户、支持任务、产品广度、能力深度、实施成本、结果证据、价格结构、分销覆盖、生态集成、服务强度、信任要求和切换成本。每个维度都需要操作定义和允许使用的证据来源。

Use two-axis maps sparingly. An axis should be independent, consistently observable, and important to the buyer or strategy. “Innovative” and “high quality” are usually weak. “Self-serve setup to assisted implementation” and “single workflow to multi-workflow platform” are easier to observe and debate with evidence.

谨慎使用二维图。坐标轴应相互独立、能够一致观察，并且对买方或战略真正重要。“创新”和“高质量”通常很弱；“自助配置到辅助实施”“单一工作流到多工作流平台”则更容易观察，也能依据证据讨论。

If the task becomes a measurable product comparison, connect the landscape to competitive benchmarking. Landscape dimensions explain structure; benchmarking metrics measure selected performance. Combining them without distinction makes subjective positioning look like test data.

如果任务变成可测量的产品比较，应连接到竞争对标。格局维度解释结构，对标指标测量选定表现。若不加区分地混合两者，会让主观定位看起来像测试数据。

Step 5: Record Sources, Dates, and Confidence步骤五：记录来源、日期和置信度

Every material cell, score, or placement should trace to evidence. Record the claim, entity, source title, URL or file, publication date, access date, market or product version, analyst, and confidence. Separate observed facts from interpretations. “The pricing page lists an annual plan” is observed; “the company is moving upmarket” is an inference requiring supporting signals.

每个重要单元格、评分或位置都应追溯到证据。记录主张、实体、来源标题、URL 或文件、发布日期、访问日期、市场或产品版本、分析人员和置信度。把观察事实与解释分开。“定价页列出年度方案”是观察；“公司正在向高端市场移动”是需要更多信号支持的推断。

Set freshness rules by field. Leadership, pricing, availability, partnerships, and product claims may change quickly; founding date may not. Never call a claim current merely because its page was accessed today. Record the date of the underlying information whenever possible.

按字段设置新鲜度规则。管理层、定价、可用性、合作关系和产品主张可能快速变化；成立日期通常不会。不能仅因今天访问了页面就称主张“最新”。应尽可能记录底层信息日期。

Step 6: Analyze Market Movement and Trends步骤六：分析市场动态与趋势

Track product launches, pricing changes, distribution agreements, standards adoption, regulatory events, hiring patterns, acquisitions, shutdowns, and repeated customer complaints. A single announcement is an event; a trend requires related observations over time and a plausible mechanism. Maintain a trend register with first and latest observation, affected segments, supporting and counter-evidence, confidence, and decision implication.

跟踪产品发布、定价变化、分销协议、标准采用、监管事件、招聘模式、并购、退出和重复客户投诉。单次公告只是事件；趋势需要随时间出现的相关观察和可信机制。趋势台账应记录首次与最近观察、受影响细分、支持与反向证据、置信度和决策影响。

Step 7: Test Scenarios and Set Decision Triggers步骤七：测试情景并设置决策触发条件

Then select two or three high-impact uncertainties such as regulation, platform entry, channel consolidation, technology cost, buyer budget, or standards adoption. Build plausible scenarios and ask how segments, positions, economics, and buyer choice change. The goal is not a perfect forecast; it is to expose fragile assumptions and find decisions that remain sensible across several futures.

然后选择两到三个影响较大的不确定因素，例如监管、平台进入、渠道整合、技术成本、买方预算或标准采用。建立可信情景，并判断细分、定位、经济模式和买方选择如何变化。目标不是完美预测，而是暴露脆弱假设，并找出在多种未来下仍然合理的决策。

Document leading indicators and triggers. “If two major channels launch private-label alternatives, reassess the partner-led entry plan” is more useful than “channel power may increase.” Triggers create a monitored decision system without pretending the data is real-time.

记录领先指标和触发条件。“如果两家主要渠道推出自有品牌替代方案，就重新评估合作伙伴主导的进入计划”，比“渠道权力可能上升”更有用。触发条件建立受监控的决策系统，同时不假装数据是实时的。

Step 8: Convert Findings into Testable Actions步骤八：把发现转化为可测试行动

Write implications as a chain: evidence, interpretation, decision relevance, action, owner, due date, and validation measure. “Three direct players offer a low-friction entry package” is evidence. “Setup effort is becoming table stakes for small teams” is interpretation. “Prototype guided onboarding and test completion with ten target users” is an action.

按照链条编写影响：证据、解释、决策相关性、行动、负责人、截止日期和验证指标。“三家直接竞品提供低门槛入门方案”是证据；“配置成本正在成为小团队的基础要求”是解释；“制作引导式上手原型并让十名目标用户测试完成率”才是行动。

Prioritize by decision value, confidence, reversibility, cost, and time to learn. Separate act now, validate next, watch, and reject items. A crowded segment is not automatically unattractive, and an empty quadrant is not automatically an opportunity. Emptiness may signal weak demand, poor economics, regulation, or a dimension buyers do not value.

根据决策价值、置信度、可逆性、成本和学习速度确定优先级。区分立即行动、下一步验证、持续观察和拒绝事项。拥挤细分不一定没有吸引力，空白象限也不一定是机会；空白可能意味着需求不足、经济性差、监管限制，或买方不重视该维度。

1. Frame: decision, owner, boundary, exclusions, proof standard.界定：决策、负责人、边界、排除项和证明标准。
2. Discover: direct, indirect, adjacent, service, internal, and non-consumption alternatives.发现：直接、间接、相邻、服务、内部和不消费替代方案。
3. Map: observable segments and buyer-relevant dimensions.映射：可观察细分和买方相关维度。
4. Verify: sources, dates, confidence, unknowns, and counter-evidence.验证：来源、日期、置信度、未知项和反证。
5. Act: implication, test, owner, deadline, metric, and review trigger.行动：影响、测试、负责人、截止日期、指标和复核触发条件。

Use AI to Organize Evidence, Not to Invent the Market用 AI 组织证据，而不是虚构市场

AI can normalize supplied player names, summarize provided documents, propose comparison fields, cluster observations, expose missing values, and draft implications for review. It is not an authoritative source or proof that a company, feature, price, or trend is current. Retain links and files, resolve entity ambiguity, verify material claims, and make a person accountable for conclusions.

AI 可以规范已提供的玩家名称、总结文档、提出比较字段、聚类观察、暴露缺失值，并起草供复核的影响。它不是权威来源，也不能证明公司、功能、价格或趋势仍然有效。应保留链接与文件，解决实体歧义，验证重要主张，并明确由人员对结论负责。

For verified input requirements and capability boundaries, see the competitor analysis tool guide. The tool does not replace customer research, audited market-share data, legal competition review, or human accountability.

有关经验证的输入要求和能力边界，请参阅竞品分析工具指南。该工具不能替代客户研究、经审计的市场份额数据、法律竞争审查或人员责任。

Worked Example: A Hypothetical Product-Evidence Market实践示例：假设的产品证据管理市场

A product team is deciding whether to build an evidence workspace for mid-market hardware companies. Its brief covers Europe, product managers and research leads, pre-launch evidence coordination, and a twelve-month horizon. Discovery reveals dedicated research repositories, general project platforms, consulting services, spreadsheets, and shared drives. The team segments them by solution form and implementation model, not by vendor category names.

某产品团队正在决定是否为中型硬件公司构建证据工作空间。简报范围涵盖欧洲、产品经理与研究负责人、发布前证据协调和十二个月时间跨度。发现过程识别出专用研究资料库、通用项目平台、咨询服务、电子表格和共享盘。团队按照解决方案形态与实施模式细分，而不是采用供应商自选品类名称。

Evidence indicates that dedicated products offer stronger traceability but require setup, general platforms have broad adoption but inconsistent research structure, and services reduce internal workload but cost more per project. Interviews suggest teams value source traceability and review ownership more than automated scoring. That interpretation remains medium confidence because the sample is small. After validating the dimensions, the team can structure a product comparison in the analyzer without treating its output as verified market evidence.

证据显示，专用产品可追溯性更强但需要配置，通用平台普及度高但研究结构不一致，服务可减少内部工作量但单项目成本更高。访谈表明，团队对来源可追溯性和复核责任的重视高于自动评分。由于样本较小，这一解释仍标记为中等置信度。验证维度后，团队可以在分析器中组织产品比较，但不能把输出当作已验证的市场证据。

Instead of declaring an “open quadrant,” the team chooses a reversible test: prototype a lightweight evidence register with explicit reviewers, recruit eight target teams from two countries, and measure setup time, completion, and willingness to replace spreadsheets. The landscape has produced an experiment, an owner, and a review date—not false certainty.

团队没有宣布存在“空白象限”，而是选择可逆测试：制作带明确复核人的轻量证据台账原型，从两个国家招募八个目标团队，测量配置时间、完成率和替代电子表格的意愿。格局最终产生实验、负责人和复核日期，而不是虚假确定性。

Common Mistakes and a Final Validation Check常见错误与最终验证检查

• Starting with famous brands: define the buyer job and inclusion rule first.从知名品牌开始：先定义买方任务和纳入规则。
• Mixing entity levels: do not compare a parent company, business unit, product, and feature as equals.混合实体层级：不要把母公司、业务单元、产品和功能当作同等实体。
• Using weak axes: choose observable, independent dimensions tied to buyer choice.使用薄弱坐标轴：选择可观察、独立且与买方选择相关的维度。
• Hiding unknowns: mark missing data and low-confidence inference instead of guessing.隐藏未知项：标明缺失数据和低置信度推断，不要猜测。
• Ending at the map: assign tests, owners, metrics, triggers, and review dates.止步于地图：分配测试、负责人、指标、触发条件和复核日期。

Frequently Asked Questions常见问题