Appendix附录

Sources, original figures and research quality notes 来源、原始图表与研究质量说明

This page holds everything needed to audit the study: the twelve references exactly as the report lists them, the report's eight original charts reproduced unaltered, all fifteen data notes with what the source says and what would be required to resolve each one, plus the publication metadata and the checks still outstanding before external release. 本页汇集审计这份研究所需的一切：12 项参考文献（按原文列出）、原报告 8 张图表（原样重现）、15 条数据说明（逐条给出原文所述与解决所需），以及发布元数据与对外发布前仍需完成的核验事项。

Report ID报告编号AUDI-A5-B10-WEU-MR-2026-11
Collection window采集区间Aug 2024 – Nov 20262024 年 8 月 – 2026 年 11 月
Discussion samples讨论样本448 verified448 条经核验
Data notes数据说明15 documented15 条已记录
Section AA 节

References [1]–[12]参考文献 [1]–[12]

The report lists twelve references and cites them inline by number. The list below reproduces each entry exactly as published — title and outlet only — and adds two things the source does not provide: the specific claim each reference supports in this build, and the section where it is cited. 原报告列出 12 项参考文献，并在正文中按编号引用。以下逐条原样重现——仅标题与媒体名——并补充两项原文未提供的信息：每项引用在本次改造中支撑的具体论断，以及它在文中被引用的位置。

Reading note阅读提示

None of the twelve entries carries a URL, publication date, author or access date, so none of them can be opened from this site. No links have been added: a guessed URL would be an invented source (DN-14). Entries are numbered and attributed exactly as the report presents them. Where a reference is a forum thread or a video, the outlet name is what the report gives — not a verified canonical location. 12 条条目均不含 URL、发布日期、作者或访问日期，因此无法从本站打开任何一条。本站未添加任何链接：猜测的 URL 等同于虚构来源（DN-14）。条目编号与归属完全依照原文。当引用为论坛帖或视频时，“媒体名”仅为原文所给，并非经核验的规范地址。

Reference [11] is a single rental experience in Spain. It is cited in the source as evidence for software behaviour across the model line; that step from one rental vehicle to a general software statement is the report's, and is flagged rather than smoothed over. 参考文献 [11] 为西班牙一次租赁体验的单一记录。原文将其用作该车型软件表现的一般性证据；这一从单台租赁车到整体软件论断的跳跃属于原报告，此处标记而不加以平滑。

Section BB 节

The eight original figures, unaltered八张原始图表（原样保留）

Every chart in the source document is reproduced here exactly as it appears in the PDF, as evidence for the rebuilt charts on the research pages. They are kept unaltered precisely because they have defects: several carry unreadable or clipped labels, raw variable names in the legend, and no data values. The numbers used on this site are taken from the accompanying tables, never read off these images. 原始文档中的每张图表都按其在 PDF 中的原样重现于此，作为研究页上重建图表的溯源证据。之所以原样保留，恰恰因为它们存在缺陷：部分图表标签不可读或被截断、图例显示原始变量名、且没有数据值。本站使用的数字均取自配套表格，绝不从这些图像中读取。

Original figure: Audi A5 Sales Volume Trend 2025-2026
Figure A1 · Audi A5 Sales Volume Trend 2025-2026图 A1 · 奥迪 A5 销量趋势 2025–2026
Area chart on a single continuous x-axis. It plots a full year (16,776) beside two half-years (8,277 / 8,882) and interpolates between them, so the falling curve reads as a continuous trend despite three unequal periods. No data labels. Corresponds to §2.1 and DN-8.单一连续横轴上的面积图。将全年度（16,776）与两个半年度（8,277 / 8,882）并列并在其间插值，因此尽管三个统计期长度不等，下行曲线仍被读作连续趋势。无数据标签。对应 §2.1 与 DN-8。
Rebuilt as discrete bars in §2.1已在 §2.1 重建为分立柱
Original figure: Western Europe Discussion Share by Country
Figure A2 · Western Europe Discussion Share by Country图 A2 · 西欧讨论份额（按国家）
Seven bars, five labels: the Germany, France and Spain labels are dropped by the generator's label-skipping rule, so three of the seven markets are unlabelled. The legend prints the raw variable name “discussion_share_pct”. No data labels. Corresponds to §2.3.七根柱、五个标签：德国、法国、西班牙三个标签被生成脚本的“隔项标注”规则省略，七个市场中有三个无标签。图例显示原始变量名“discussion_share_pct”。无数据标签。对应 §2.3。
Rebuilt with all seven labels in §2.3已在 §2.3 重建并补齐七个标签
Original figure: Audi A5 Discussion Channel Distribution
Figure A3 · Audi A5 Discussion Channel Distribution图 A3 · 奥迪 A5 讨论渠道分布
Pie chart with clipped callout labels — “Official Corporate Med…”, “Video Content Platforms:19 (1…”, “Dealer & Retail We…”. Label format is also inconsistent: one entry reads “28 (28%)”, others are truncated before the percentage. Corresponds to §2.4.饼图的引线标签被截断——“Official Corporate Med…”“Video Content Platforms:19 (1…”“Dealer & Retail We…”。标签格式也不统一：一项写作“28 (28%)”，其余在百分比之前即被截断。对应 §2.4。
Rebuilt as a donut in §2.4已在 §2.4 重建为环形图
Original figure: Positive Feedback Volume by Category
Figure A4 · Positive Feedback Volume by Category图 A4 · 分类正面反馈量
Cleanest of the eight: four labelled bars on a 0–80 axis, legend “mentions”. It carries no value labels and no base, so it conveys rank only — the published percentages (47.9 / 32.5 / 15.3 / 4.3) cannot be read from it. Corresponds to §3.八张中最为整洁的一张：四根带标签的柱、纵轴 0–80、图例“mentions”。但无数据标签、无基数，因此只能传达排序——已发布的百分比（47.9 / 32.5 / 15.3 / 4.3）无法从此图读出。对应 §3。
Rebuilt with counts and shares in §3已在 §3 以次数与占比重建
Original figure: Neutral Feedback Category Distribution
Figure A5 · Neutral Feedback Category Distribution图 A5 · 分类中性反馈分布
Pie chart in the same sequence as the table (31 / 24 / 22 / 23). Two of four labels are clipped (“Pricing & Trim Configu…”, “Powetrain Specificati…”) — note the misspelling inside the image itself. Corresponds to §4 and DN-3.饼图顺序与表格一致（31 / 24 / 22 / 23）。四个标签中有两个被截断（“Pricing & Trim Configu…”“Powetrain Specificati…”）——注意图像内部本身即存在拼写错误。对应 §4 与 DN-3。
Rebuilt as a donut in §4已在 §4 重建为环形图
Original figure: Negative Feedback Category Distribution
Figure A6 · Negative Feedback Category Distribution for Audi A5 Western Euro…图 A6 · 奥迪 A5 西欧分类负面反馈分布
The title itself is clipped at the right edge. Six bars, three labels — Pricing, Software and Other are unlabelled, which means the second-largest category is invisible in the chart while its bar remains. Legend “share_percent”. Corresponds to §5 and DN-4.标题本身在右缘被截断。六根柱、三个标签——价格、软件与“其他”没有标签，这意味着第二大类别在图中不可见，而其柱体仍在。图例“share_percent”。对应 §5 与 DN-4。
Rebuilt with all six labels in §5已在 §5 重建并补齐六个标签
Original figure: Competitive Dimension Performance Rating
Figure A7 · Competitive Dimension Performance Rating图 A7 · 竞品维度表现评分
Grouped bars for six dimensions, but only three axis labels survive (Exterior Design, Premium Quality, Pricing), so half the dimensions are unlabelled. Legend uses machine variable names: audi_a5_rating, bmw_3_rating, mercedes_c_rating, volvo_v60_rating. Corresponds to §7.2.六个维度的分组柱，但仅保留三个横轴标签（外观设计、高级质感、价格），一半维度无标签。图例使用机器变量名：audi_a5_rating、bmw_3_rating、mercedes_c_rating、volvo_v60_rating。对应 §7.2。
Rebuilt with six dimensions and brand names in §7.2已在 §7.2 重建并补齐六维度与品牌名
Original figure: Factor Correlation with Market Reach
Figure A8 · Factor Correlation with Market Reach图 A8 · 因子与市场覆盖的相关性
Five bars, three labels (Infotainment Usability, Naming Confusion, Hybrid Powertrain) — Pricing and Exterior Design are unlabelled. All bars share one colour, so the source's HIGH / MEDIUM / LOW classification cannot be read from the chart. Legend “correlation_score”. Corresponds to §8.3.五根柱、三个标签（座舱易用性、命名混淆、混动总成）——价格与外观设计无标签。所有柱体同一颜色，因此原文的 HIGH / MEDIUM / LOW 分级无法从图中读出。图例“correlation_score”。对应 §8.3。
Rebuilt with labels and impact bands in §8.3已在 §8.3 重建并补齐标签与影响等级

The images above are the PDF's embedded bitmaps at their original 800 × 600 resolution, extracted without rescaling or retouching. Their defects are a property of the source document, not of this site. Any figure value quoted anywhere on this site traces to a table, and those tables are reproduced with a cumulative column so that the arithmetic can be checked. 以上图像为 PDF 内嵌位图的原始 800 × 600 分辨率版本，未经缩放或修饰地提取。其缺陷属于原始文档的属性，而非本站所为。本站任何位置引用的图表数值均可追溯至表格，且相关表格均附累计列以便核对算术。

Section CC 节

Research quality notes — DN-1 to DN-15研究质量说明——DN-1 至 DN-15

The source report carries a heading called “Research Overview and Data Notes” but publishes no data notes beneath it (DN-15). These fifteen notes were written for this website: each records what the source states, why the issue matters commercially, and what would be required to resolve it. None of them changes a source figure, and none supplies a missing denominator. 原始报告设有“Research Overview and Data Notes”标题，但标题之下未刊出任何数据说明（DN-15）。以下 15 条说明为本站撰写：每一条记录原文所述、该问题在商业上为何重要，以及解决它需要什么。任何一条都未改动原文数字，也未替原文补齐缺失的分母。

  1. Sentiment shares have no declared base情绪份额未声明基数

    What the source states: positive 38%, neutral 22%, negative 40% of “total classified user feedback”, described elsewhere as 448 verified discussion samples.原文所述：正面 38%、中性 22%、负面 40%，口径为“已分类用户反馈总量”，而另一处称样本为 448 条经核验讨论。

    Why it matters: the three figures total exactly 100%, so they describe the composition of a classified set — not of the 448 samples. Posts that were irrelevant, duplicate or unclassifiable are invisible in that denominator. A reader who assumes “40% of 448 discussants were negative” is reading a number the report never published.为何重要：三个数字恰好合计 100%，说明它们描述的是一个“已分类集合”的构成，而非 448 条样本的构成。无关、重复或无法分类的帖文在这个分母中不可见。若读者认为“448 位讨论者中 40% 持负面”，那是在读一个报告从未发布的数字。

    To resolve: publish the classification funnel — samples collected, samples excluded and why, samples classified, and the resulting denominator.解决所需：公布分类漏斗——采集样本数、排除样本数及原因、完成分类样本数，以及由此得出的分母。

    Applies to §3, §4, §5 and every sentiment figure on the Executive Overview适用：§3、§4、§5 及高层概览上的全部情绪指标
  2. Positive mention counts do not reconcile to the sample正面提及数与样本不对账

    What the source states: 78 + 53 + 25 + 7 = 163 positive mentions, with shares of 47.9 / 32.5 / 15.3 / 4.3% summing to 100.0%.原文所述：78 + 53 + 25 + 7 = 163 次正面提及，占比 47.9 / 32.5 / 15.3 / 4.3%，合计 100.0%。

    Why it matters: 47.9% of 163 is 78 — the arithmetic closes within the mention base. But 163 mentions against 448 samples implies roughly 2.75 coded mentions per sample with no stated multi-coding rule, and the positive-discussion base is never given. The counts are therefore usable as a ranking of attributes, not as a measure of how many buyers raised each one.为何重要：163 的 47.9% 正好是 78——算术在“提及基数”内闭合。但以 163 次提及对应 448 条样本，意味着每条样本平均产生约 2.75 次编码提及，而原文未说明多重编码规则，也未给出正面讨论基数。因此这些次数可用于属性排序，不能用于衡量有多少买家提及某项。

    To resolve: state the multi-coding rule and the positive-only base, or publish per-sample coding so the 163 can be rebuilt.解决所需：说明多重编码规则与“仅正面”基数，或公布逐条样本的编码以便重建 163。

    Applies to §3 and Table 3.1适用：§3 与表 3.1
  3. Neutral category order is not a ranking中性类别顺序不构成排序

    What the source states: pricing 31%, launch timing 24%, powertrain 22%, body style 23% — in that order, in both the table and the figure.原文所述：价格 31%、上市时间 24%、动力总成 22%、车身形式 23%——表格与图表均按此顺序。

    Why it matters: the sequence is not descending, so 23% is printed after 22%. A reader scanning the list takes the order as implied priority, and in this case the order contradicts the values. The totals do reconcile to 100%, so the only issue is presentational sequence.为何重要：该顺序并非降序，因此 23% 排在 22% 之后。读者扫读列表时会把顺序当作隐含的优先级，而在此情形下顺序与数值相矛盾。合计确实闭合至 100%，因此唯一的问题是呈现顺序。

    To resolve: sort by value, or keep the source order and state that the sequence is not a ranking — which is what this site does, in both its table and its chart.解决所需：按数值排序，或保留原顺序并声明该顺序不是排序——本站采用后者，表格与图表一致。

    Applies to §4 and Table 4.1适用：§4 与表 4.1
  4. Negative shares total 98%, and “Other” is the second-largest line负面份额合计 98%，“其他”为第二大项

    What the source states: interior interface 27%, pricing 22%, naming 11%, software 8%, driving dynamics 7%, other 25% — as printed, these sum to 98%.原文所述：内饰交互 27%、价格 22%、命名 11%、软件 8%、驾驶动态 7%、其他 25%——按原文印刷，合计为 98%。

    Why it matters: two separate problems. First, 2% is unallocated in a table presented as a share of negative feedback. Second, “Other” at 25% is larger than every named category except interior interface, so the five named pain points explain at most 73% of negative sentiment — and the largest single block of criticism is simply uncharacterised. Any statement that the listed issues account for negative sentiment is therefore overstated.为何重要：两个独立问题。其一，在一张“占负面反馈”的表中，有 2% 未归集。其二，“其他”高达 25%，除内饰交互外大于任何具名类别，因此五个具名痛点至多解释负面情绪的 73%——而最大的一块批评本身完全没有刻画。任何“所列问题即构成负面情绪”的表述因此都是高估。

    To resolve: publish the “Other” sub-categories and the disposition of the residual 2%. This site retains both as published and does not redistribute them.解决所需：公布“其他”的子类别以及 2% 残差的处置方式。本站原样保留两者，未作重新分配。

    Applies to §5 and Table 5.1适用：§5 与表 5.1
  5. 35% “practical design execution” versus 32.5% Avant practicality35% “实用设计执行”与 32.5% “Avant 实用性”

    What the source states: the sentiment synthesis attributes 35% of all positive feedback to practical design execution; the positive table gives Avant practicality 32.5%.原文所述：情绪综合章节将全部正面反馈的 35% 归因于“实用设计执行”；正面反馈表给出 Avant 实用性 32.5%。

    Why it matters: the synthesis figure is larger than the only positive category that matches it, which suggests the 35% aggregates Avant practicality with liftback Sedan utility comments. That is a plausible reading — but it is a reading, not a published definition, and it means the two numbers cannot be used interchangeably across the site.为何重要：综合章节的数字大于与之唯一对应的正面类别，暗示 35% 是把 Avant 实用性与掀背 Sedan 实用性评论合并而来。这一解读可以成立——但它是解读，不是原文定义，这意味着两个数字不能跨站互替使用。

    To resolve: restate the 35% as a compound category with its components, or reconcile it to the 32.5%.解决所需：将 35% 重新表述为一个复合类别并列出其构成，或与 32.5% 对账。

    Applies to §3.4 and §8.1适用：§3.4 与 §8.1
  6. The pricing arithmetic cannot be recomputed价格算术无法复算

    What the source states: base Sedan €47,900–€49,300; Avant consistently +€2,100; optioned cars regularly above €65,000 before delivery; 11–14% above the outgoing A4 Avant at launch; 4–7% entry premium over direct competitors.原文所述：Sedan 基础价 47,900–49,300 欧元；Avant 恒定 +2,100 欧元；选装车常超 65,000 欧元（不含交付）；较旧款 A4 Avant 上市时高 11–14%；相对直接竞品入门价高 4–7%。

    Why it matters: no outgoing price and no competitor price is published, so not one of these relationships can be checked by calculation. The 11–14% comparison also sets the new Sedan's base against the outgoing Avant base — different body styles, which alone biases the percentage. And 4–7% and 11–14% are two different gaps used in two different arguments; they are not interchangeable.为何重要：原文既未公布旧款价格，也未公布竞品价格，因此没有任何一项关系可以通过计算检验。此外，11–14% 的比较是以新款 Sedan 基础价对照旧款 Avant 基础价——车身形式不同，本身就使该百分比产生偏差。而 4–7% 与 11–14% 是用于两处不同论证的两个不同价差，不可互换。

    To resolve: publish the four input prices with their model year, body style and market, then recompute both percentages.解决所需：公布四个输入价格及其年款、车身形式与市场，再重算两个百分比。

    Applies to §5.2, §8.3 and the price KPI on the Executive Overview适用：§5.2、§8.3 以及高层概览的价格指标
  7. Two market denominators that must not be merged两个不可合并的市场分母

    What the source states: Germany accounts for 29% of Audi brand deliveries in Q1 2026 [2]; Germany accounts for 28% of A5 discussion volume.原文所述：2026 年第一季度德国占奥迪品牌交付量 29% [2]；德国占 A5 讨论量的 28%。

    Why it matters: the two figures are close and describe different things — one is a share of deliveries for the whole brand in one quarter, the other a share of discussion volume for one model line across the full collection window. Adding, averaging or presenting them as “Germany's share” would create a number that exists nowhere in the source.为何重要：两个数字接近，但描述的对象不同——一个是单季度内整个品牌的交付份额，另一个是单一车型线在全采集区间内的讨论量份额。相加、取平均或表述为“德国份额”，都会产生一个原文中并不存在的数字。

    To resolve: keep both, labelled with their own denominator, as this site does in Table 2.2.解决所需：两者并列保留并各自标明分母，本站表 2.2 即按此处理。

    Applies to §2.3 and the geographic KPI适用：§2.3 与地理指标
  8. One sequence, three unequal periods同一序列，三个不等长区间

    What the source states: the section is headed “Discussion Volume Trend”, while its figure and table present unit sales: FY2025 16,776, H1 2025 8,277, H1 2026 8,882, and the figure joins them with a smoothed continuous curve.原文所述：该小节标题为“Discussion Volume Trend”，而其图表与表格呈现的是销量：FY2025 16,776 台、H1 2025 8,277 台、H1 2026 8,882 台，且图表以平滑连续曲线将它们相连。

    Why it matters: a twelve-month period and two six-month periods cannot share a continuous time axis; drawn that way, the “trend” from 16,776 down to 8,277 is an artefact of period length, and the −32% / +7% year-on-year comparisons are the only defensible readings. The heading also invites the data to be read as discussion volume when it is sales volume.为何重要：十二个月的区间与两个六个月的区间无法共享一条连续时间轴；按此绘制，从 16,776 降至 8,277 的“趋势”只是区间长度的产物，唯一可成立的读法是 −32% 与 +7% 的同比。标题还会诱导读者把销量数据读作讨论量。

    To resolve: plot the three periods as discrete bars, label each with its period length, and separate sales volume from discussion volume. This site does exactly that.解决所需：将三个区间绘为分立柱、逐柱标注区间长度，并把销量与讨论量分开呈现。本站已按此处理。

    Applies to §2.1 and the sales chart适用：§2.1 与销量图
  9. The analysis window outruns the sales data分析区间长于销量数据覆盖

    What the source states: a collection window of August 2024 to November 2026 (28 months), sales data only through H1 2026, and a software statement that issues remain “18 months after launch”.原文所述：采集区间为 2024 年 8 月至 2026 年 11 月（28 个月），销量数据仅到 2026 年上半年，并有“上市 18 个月后问题仍未解决”的软件表述。

    Why it matters: three different clocks run through the report — discourse, sales and product age — and they do not close. Sales-based conclusions cannot be extended to the final months of the collection window, and “18 months” cannot be dated precisely without a launch date the report never gives.为何重要：报告中同时存在三个不同的时间基准——话语、销量与产品年龄——彼此并不闭合。基于销量的结论不能外推到采集区间的最后几个月；而“18 个月”在原文未给上市日期的前提下无法精确对应到日历日期。

    To resolve: state each dataset's coverage period separately, and date the launch reference point.解决所需：分别标明每个数据集的覆盖区间，并给出上市基准日期。

    Applies to §2.2, §5.4 and the report header适用：§2.2、§5.4 与报告页头
  10. Undeclared subset denominators: 78%, 72%, 68%未声明的子集分母：78%、72%、68%

    What the source states: 78% of configuration inquiries concern driver assistance, regional equipment and the liftback hatch; 72% of forum posts mentioning the model name express naming confusion; 68% of timing posts are neutral status updates.原文所述：78% 的配置询问涉及驾驶辅助、区域装备与掀背开口；提及车型名的论坛帖中 72% 表达命名困惑；时间类帖文中 68% 为中性状态更新。

    Why it matters: each is a share of its own unspecified subset — configuration questions, name-mentioning posts, timing posts. None can be converted into a share of 448, and none can be compared with another. They also carry no market split, so “German delivery queries” cannot be given a rate, only a rank.为何重要：每个数字都是各自未指定子集内的占比——配置类提问、提及车型名的帖文、时间类帖文。均不能换算为 448 条样本中的比例，彼此也不能比较。它们也没有分市场拆分，因此“德国交付询问”只能给出相对高低，不能给出比例。

    To resolve: publish the subset size alongside each percentage, and add market-level splits.解决所需：在每个百分比旁公布其子集规模，并补充分市场拆分。

    Applies to §4.1, §4.2 and §5.3适用：§4.1、§4.2 与 §5.3
  11. Competitive ratings: method not disclosed竞品评分：方法未公开

    What the source states: a ten-point rating for four brands across six dimensions — for example exterior design 8.7 / 8.1 / 8.3 / 7.9, pricing 6.9 / 7.8 / 7.5 / 7.2.原文所述：四品牌六维度的十分制评分——例如外观设计 8.7 / 8.1 / 8.3 / 7.9，价格 6.9 / 7.8 / 7.5 / 7.2。

    Why it matters: no rater, panel size, aggregation rule or scale anchors are given, and it is not stated whether the scores reflect media assessment or user perception. A 0.2-point gap between two brands is therefore below any defensible resolution, yet the table invites exactly that reading. Scores are reproduced as report-stated assessments.为何重要：原文未给出评分人、评审人数、聚合规则或量表锚点，也未说明评分反映媒体评价还是用户感知。因此两品牌之间 0.2 分的差距低于任何可辩护的分辨率，而表格恰恰诱导这种读法。本站按“原文所述评估”原样呈现。

    To resolve: publish the rating protocol, and report at the precision the method supports — one decimal place implies a resolution the method does not demonstrate.解决所需：公布评分方法，并按方法所能支撑的精度报告——保留一位小数意味着一种方法并未证明的分辨率。

    Applies to §7.2, Table 7.1 and Table 6.1适用：§7.2、表 7.1 与表 6.1
  12. Cross-shopping overlap: no sample, no window, overlapping concepts交叉考虑重叠：无样本、无窗口、概念重叠

    What the source states: cross-consideration of 68% with BMW, 59% with Mercedes-Benz, 31% with Volvo, and regional alternatives below 12%.原文所述：与宝马交叉考虑 68%、奔驰 59%、沃尔沃 31%，区域替代品低于 12%。

    Why it matters: no sample, survey window, geography or definition of “cross-shopping activity” is given. The three rates overlap conceptually — one buyer can consider two or three rivals — so they cannot be summed, averaged or read as a partition of demand. The sub-12% regional figure has the same defect.为何重要：原文未给样本、调研窗口、地域范围或“交叉考虑活动”的定义。三个比率在概念上重叠——同一买家可能考虑两到三个竞品——因此不可相加、取平均，也不能读作需求的分割。区域替代品低于 12% 的表述存在同样问题。

    To resolve: publish the base, and present the rates as multi-select shares with the question wording.解决所需：公布基数，并按多选份额呈现这些比率，同时给出提问原文。

    Applies to §7.1 and §7.3适用：§7.1 与 §7.3
  13. Correlations: arithmetic consistent, direction unstated相关性：算术自洽，方向未说明

    What the source states: infotainment usability correlates 0.82 with market reach and explains 67% of variance; pricing correlates 0.76; naming 0.61, exterior design 0.57, hybrid powertrain 0.43, classified HIGH / MEDIUM / LOW.原文所述：座舱易用性与市场覆盖相关系数 0.82、解释 67% 方差；价格 0.76；命名 0.61、外观设计 0.57、混动总成 0.43，并分为 HIGH / MEDIUM / LOW。

    Why it matters: 0.82² = 0.672, so the variance statement checks out. But no sample size, window, method or confidence interval is given, and no causal direction is asserted — a market that circulates infotainment complaints may simply be a market with more demanding buyers. A correlation of this size across aggregated market data, without a stated n, is a hypothesis rather than a measured relationship.为何重要：0.82² = 0.672，方差解释的算术成立。但原文未给样本量、窗口、方法或置信区间，也未断言因果方向——一个流传座舱抱怨的市场，也可能只是一个买家要求更高的市场。在未声明样本量的聚合市场数据上得到此种量级的相关系数，属于假设，而非已测量的关系。

    To resolve: publish n, the aggregation unit (market? quarter?), the method, and a directional hypothesis stated as a hypothesis.解决所需：公布样本量、聚合单元（市场？季度？）、方法，以及以假设形式表述的方向性判断。

    Applies to §8.2, §8.3 and the correlation KPI适用：§8.2、§8.3 与相关性指标
  14. References carry titles and outlets only参考文献仅有标题与媒体名

    What the source states: twelve numbered references, each a title plus an outlet name — for instance “Audi A5 Sales Figures | GCBC”. No URL, date, author or access date appears anywhere.原文所述：12 项编号引用，每项为标题加媒体名——例如“Audi A5 Sales Figures | GCBC”。全文未出现 URL、日期、作者或访问日期。

    Why it matters: no claim in the report can be traced to a document without manual search, and several references are community sources (a Reddit thread, two forum threads) cited alongside trade press without type distinction. The report is therefore not independently auditable as published.为何重要：不经人工检索，报告中任何论断都无法追溯到具体文档；且有数项引用为社区来源（一条 Reddit 帖、两条论坛帖），与行业媒体并列引用而未作类型区分。因此按发布状态，该报告不具备独立可审计性。

    To resolve: supply a full citation per entry — URL, date, author — and mark source type. This site supplies no links, because a guessed URL would be an invented source.解决所需：为每条引用提供完整信息——URL、日期、作者——并标注来源类型。本站不提供任何链接，因为猜测的 URL 等同于虚构来源。

    Applies to the reference list above and every inline [n] citation适用：上方参考文献列表与全文所有 [n] 引用
  15. The “Data Notes” heading is empty“Data Notes”标题下无内容

    What the source states: the methodology section is headed “Research Overview and Data Notes”, but the section contains no data notes — only method and scope statements.原文所述：方法章节标题为“Research Overview and Data Notes”，但该节内并无任何数据说明——只有方法与范围陈述。

    Why it matters: the heading promises exactly the disclosure that the report most needs and does not deliver: denominators, subset sizes, rating protocol, correlation method. A reader who trusts the heading will assume the caveats have been handled.为何重要：该标题承诺的恰恰是报告最需要披露却未披露的内容：分母、子集规模、评分方法、相关性方法。信任该标题的读者会以为这些限定已被处理。

    To resolve: populate the section, or retitle it. This site's fifteen notes are an attempt to fill that gap without altering the report's own figures.解决所需：补写该节内容，或修改标题。本站的 15 条说明正是为填补这一空缺所作的努力，且未改动报告自身数字。

    Applies to §1 and to this page适用：§1 与本页
Section DD 节

Publication metadata发布元数据

A single record in the structure required for future consolidation into a multi-report portal. The same values are published machine-readable in report.json and summarised in reports-index.json. Fields the source does not establish are left empty rather than inferred. 本报告在未来并入多报告门户时所需结构中的单条记录。相同值以机器可读形式发布于 report.json，并在 reports-index.json 中摘要呈现。原文未能确认的字段一律留空，不作推断。

reportIdAUDI-A5-B10-WEU-MR-2026-11
slugaudi-a5-b10-weu — intended route /reports/audi-a5-b10-weu/
brand品牌Audi
parentGroup所属集团Volkswagen Group
model车型A5 Sedan & Avant (B10 generation)
industry行业Automotive — premium mid-size passenger cars汽车——高端中型乘用车
geographicScope地域范围Western Europe — Italy, Germany, United Kingdom, France, Netherlands, Spain, Belgium西欧——意大利、德国、英国、法国、荷兰、西班牙、比利时
researchPeriod研究区间2024-08 → 2026-11
reportDate报告日期2026-11-09
availableLanguages可用语言en · zh-CN
sample样本448 verified public discussion samples448 条经核验公开讨论样本
sourceDocument原始文档audi-a5-b10-weu-source-report.pdf — 14 pages
publicationStatus发布状态internal review draft内部审阅稿 not cleared for external release — see Section F尚未通过对外发布审核——见 F 节
reportOwner报告负责人To be configured before external release对外发布前需配置
contactEmail联系邮箱To be configured before external release对外发布前需配置

Contact fields are intentionally blank. The site renders a placeholder rather than an invented address, and no form submission behaviour is wired to a server. 联系方式字段刻意留空。站点渲染占位文案而非虚构地址，且未将任何表单提交行为接到服务器。

Section EE 节

What is claimed, what was checked何为主张，何为已核验

The report states findings; this site rebuilt them faithfully and checked only what can be checked without external sources — internal consistency and arithmetic. Nothing here has been independently verified against the cited outlets, because the references carry no locatable address. 报告陈述发现；本站忠实重建，并只核验了在不依赖外部来源的情况下可核验的部分——内部一致性与算术。此处没有任何内容针对所引媒体做过独立核验，因为参考文献没有可定位的地址。

Table E1 · Verification status by claim表 E1 · 各论断的核验状态
Claim in the report报告中的论断 Check performed in this build本次改造所做的核验 Result结果
Sentiment split 38 / 22 / 40情绪分布 38 / 22 / 40 Sum check合计校验 Totals 100% of the classified set; denominator relative to 448 not published (DN-1)合计为已分类集合的 100%；相对 448 的分母未公布（DN-1）
Positive 78 / 53 / 25 / 7, shares 47.9 / 32.5 / 15.3 / 4.3正面 78 / 53 / 25 / 7，占比 47.9 / 32.5 / 15.3 / 4.3 Reciprocal arithmetic and sum check反算与合计校验 Shares reproduce from the counts and total 100%; count base does not reconcile to 448 (DN-2)占比可由次数复算且合计 100%；次数基数与 448 未对账（DN-2）
Neutral 31 / 24 / 22 / 23中性 31 / 24 / 22 / 23 Sum and ordering check合计与顺序校验 Totals 100%; sequence not descending (DN-3)合计 100%；顺序非降序（DN-3）
Negative 27 / 22 / 11 / 8 / 7 / 25负面 27 / 22 / 11 / 8 / 7 / 25 Sum check合计校验 Totals 98%, not 100%; 2% unallocated (DN-4)合计 98%，非 100%；2% 未归集（DN-4）
Country shares 41 / 28 / 14 / 7 / 5 / 3 / 2国家份额 41 / 28 / 14 / 7 / 5 / 3 / 2 Sum check合计校验 Totals 100% (DN-7 applies to the 29% brand-delivery figure, not this split)合计 100%（DN-7 针对 29% 的品牌交付数字，不针对此分布）
Channel mix 32 / 28 / 21 / 19渠道构成 32 / 28 / 21 / 19 Sum check合计校验 Totals 100%合计 100%
Sales 16,776 (−32%) / 8,277 / 8,882 (+7%)销量 16,776（−32%）/ 8,277 / 8,882（+7%） Year-on-year arithmetic同比算术 Period lengths are unequal, so no sum or trend is valid; YoY percentages are the report's and unverified against [1]区间长度不等，任何合计或趋势均不成立；同比百分比为原文所述，未对 [1] 核验
Price ladder €47,900–49,300, +€2,100, >€65,000, 11–14%, 4–7%价格阶梯 47,900–49,300、+2,100、>65,000、11–14%、4–7% Recomputation attempt复算尝试 Cannot be reproduced — no base prices published (DN-6)无法复现——未公布基准价格（DN-6）
Competitive ratings, six dimensions × four brands竞品评分，六维度 × 四品牌 Row-maximum check行内最高值校验 Highest score in each row identifiable; method not disclosed (DN-11)各行最高分可辨识；方法未公开（DN-11）
Correlations 0.82 / 0.76 / 0.61 / 0.57 / 0.43相关系数 0.82 / 0.76 / 0.61 / 0.57 / 0.43 Variance arithmetic方差算术 0.82² = 0.672 matches the stated 67%; n and method absent (DN-13)0.82² = 0.672，与所述 67% 相符；样本量、计算方法与因果方向均未给出（DN-13）
Cross-shopping 68 / 59 / 31交叉考虑 68 / 59 / 31 Partition check分割校验 Cannot be a partition — the three overlap and exceed 100% (DN-12)不构成分割——三者重叠且合计超过 100%（DN-12）
Audi acknowledged the naming error; possible 2029 reversal奥迪承认命名失误；或于 2029 年反转 Source-type review来源类型审查 Rests on two trade articles plus a Reddit thread; not verified against a primary Audi statement依据两篇行业文章加一条 Reddit 帖；未对奥迪一手声明核验
“18 months of OTA updates unresolved”“18 个月 OTA 更新仍未解决” Date-reasoning check日期推理校验 No launch date published; cannot be dated (DN-9)未公布上市日期；无法对应到具体日期（DN-9）
Residual-value consistency as an advantage保值率一致性作为优势 Evidence check证据校验 Asserted qualitatively; no residual-value data published anywhere in the report仅为定性主张；报告全文未发布任何保值率数据

“Unverified” is the correct status for every external claim in this build. The checks above test internal coherence only; they do not confirm that the underlying market facts are true. 在本次改造中，所有外部论断的正确状态都是“未核验”。上表仅检验内部自洽性，并不确认其背后的市场事实为真。

Section FF 节

Publication status and release requirements发布状态与发布要求

F.1 StatusF.1 状态

Internal review draft. The site is complete and functional as a research artefact, but it is not cleared for external distribution. The blockers are rights and provenance, not presentation. 内部审阅稿。作为研究类交付物，站点功能与内容已完整，但尚未通过对外分发审核。待解决的阻碍属于版权与溯源范畴，而非呈现问题。

F.2 What can be stated openlyF.2 可公开陈述的内容

F.3 Must be resolved before external releaseF.3 对外发布前必须解决

  1. Vehicle imagery rights. The three vehicle photographs on the overview and research pages are editorial reference images drawn from public media sources, not official manufacturer assets. Licensing must be confirmed or the images replaced with owned or licensed material before any public launch; captions already state the position.车型影像权利。概览与研究页上的三张车型照片为取自公开媒体的编辑参考图，并非厂商官方素材。公开发布前须确认授权，或替换为自有／已授权素材；图注已就此作出说明。
  2. Reference locators. Twelve citations currently carry titles and outlets only (DN-14). Each needs a URL, date and author, with source type marked, before the report can be audited by a third party.引文定位信息。12 项引用目前仅有标题与媒体名（DN-14）。在第三方能够审计之前，每条都需补 URL、日期与作者，并标注来源类型。
  3. Sample provenance. The construction of the 448-sample set — collection method, platforms, language handling, inclusion rules, coding scheme — is undocumented in the source and must be supplied before the sentiment shares are quoted externally.样本溯源。448 条样本集的构建方式——采集方法、平台、语言处理、纳入规则、编码方案——原文未记录；在对外引用情绪份额前必须补齐。
  4. Attribution of Audi statements. The naming acknowledgement and the possible 2029 reversal rest on trade coverage and a community thread. Direct quotation of a manufacturer position requires a primary source.奥迪表态的归属。命名认错与 2029 年可能反转的依据为行业报道及社区帖。直接引用厂商立场需要一手来源。
  5. Contact configuration. Owner, address and request-research endpoint must be filled in assets/js/config.js. Until then the site displays placeholders and no form is active.联系方式配置。负责人、邮箱与“定制研究”入口须在 assets/js/config.js 中填入。在此之前站点显示占位文案且无可用表单。
  6. Legal review. The disclaimer stating no affiliation with Audi AG or Volkswagen Group, and the caution that the analysis is not representative of all consumers, must be reviewed by whoever owns external communications.法务审核。关于与奥迪股份公司或大众集团无隶属关系的免责声明，以及“分析不代表全体消费者”的提示，须由对外传播负责人审核。

F.4 Rebuilding notes for maintainersF.4 给维护者的重建说明

↑