The country table (p.2) reports Spain at +79% share-of-search growth; the bar chart printed beside it plots only the United Kingdom, France and Germany.

Spain is the second-highest growth market in the table and the leading market by registrations (3,491) and discussion index (100). Read from the chart alone, the report understates its strongest market.

Chart rebuilt from tabulated values on this site; all six markets plotted, Benelux shown as N/A.

Pie-chart labels print as “Value for Money:4…”, “BYD Technology Acceptance:17…”, “Range/charging/fu…”, “Waiting list questions:24 (2…”.

A reader cannot recover the value from the chart; the label, not the number, was clipped.

Underlying tables (pp.4, 6, 7) carry complete values; charts re-rendered with full labels.

Chart legend “discussion_share” (p.9), “mention_count” (p.10); axis categories “price_value”, “brand_trust”, “delivery_wait_time”.

Reader-facing charts expose database column names, which weakens the publication’s credibility at exactly the moments it is making its strongest claims.

Labels replaced with the source table’s own wording; no value changed.

The table (p.10) lists five factors — 19, 19, 7, 5, 3. The chart prints a sixth bar labelled delivery_wait_time, roughly one mention.

No row, no count and no body-text discussion support that category. Either a row was dropped from the table or a stray series was plotted.

Sixth category not plotted on this site; table reproduced exactly as published.

The stated sample is 54 undecided buyers; the five factor counts sum to 53.

Small, but it sits in the table that carries the report’s central “dual threshold” finding, and readers will check the arithmetic.

Reported as printed; discrepancy flagged, not corrected.

Body text (p.2): “Italy recorded 400% year-on-year sales growth.” Table (p.2): Italy +53% share-of-search growth YoY.

Different metrics with different denominators. Not a contradiction — but the two must never be merged into one “growth” figure.

Kept in separate columns and separate sentences throughout this site.

Body text (p.2): “700% sales growth paired with the lowest awareness growth, indicating sales are currently driven by fleet channels rather than organic consumer demand [4].” Reference [4] is titled “BYD Triples European Sales in December, Surpassing 26,000 Vehicles.”

The cited source is a European-level December statement, not a Germany-specific ATTO 2 registration count; the fleet-channel reading is an inference with no measured channel split behind it.

Presented as the report’s claim, with this caveat attached at the point of use.

Table (p.2): Spain 3,491 and Italy 3,347; United Kingdom, France, Germany and Benelux are N/A.

Any sales-per-unit-of-attention comparison across markets is impossible with four markets missing, even though the text discusses German and Italian sales growth.

Not computed. Registration chart shows N/A explicitly.

Column header “Share of Search Growth YoY”, values +130% to +19%.

Unclear whether this is growth in share of search interest, growth in search volume, or growth in the brand’s share of segment searches. The unit (%) is explicit; the base is not.

Reproduced with the source’s own column label; no interpretation added.

Values 92, 100, 38, 87, 41, 29 with Spain = 100.

It is an index, not a mention count made comparable across markets. Plotting it beside counts, or reading “100” as “100 mentions”, would be a category error.

Plotted in its own chart, labelled an index, never combined with counts or shares.

Positive themes are a “share of positive comments”; neutral and negative themes are a “relative frequency” that the source never defines, though both sum to 100.

The three sets cannot be combined into one sentiment composition, and “relative frequency” cannot be rescaled onto a share-of-all-comments basis without inventing a denominator.

Each distribution is shown in its own chart; no combined sentiment chart exists on this site.

One table carries a percentage share, a qualitative label (“Neutral / Mixed”), an ordinal rating (“Class leading”) and an approximate distance (“~360km”).

Only two of the four rows are quantitative, and one of those is approximate. The table reads as a scorecard but is not a comparable index.

Share is charted; the remaining three rows are shown as text values only.

“35% of neutral discussion covers powertrain indecision and trim selection. 32% of prospective ATTO 2 buyers are unable to choose between BEV and DM-i variants.”

The 32% is measured on a different population (“prospective buyers”) than the 35% (“neutral discussion”), yet it is nested inside it as if it were a sub-share.

Two percentages presented separately; never multiplied or summed.

Efficiency reported as 55mpg and 3.6mi/kWh (p.4); ranges in km (p.8); prices in € (pp.4, 9); and the hybrid/PHEV row states “€9,000 lower entry price” without naming the rival.

Imperial efficiency figures are unhelpful for a European audience, and a price gap without a stated comparator cannot be checked.

No conversion and no comparator added — either would introduce numbers the report does not publish.

The geographic scope is stated as Western Europe; the six markets (UK, Spain, France, Italy, Germany, Benelux) appear only in the country table, and Benelux is treated as one market.

Scope determines what the findings can be generalised to. A reader cannot tell whether Nordics, Austria or Portugal were in scope and simply absent, or out of scope.

Scope described on this site exactly as the source states it, with the market list attributed to the country table.

Opportunity segment table (p.11) gives market share and success probability for four segments (41/82, 27/76, 19/61, 13/29).

No definition, method, sample or confidence basis is given. It reads like a forecast but carries no forecast apparatus.

Presented as a directional ranking, never as a probability or a projection.

Priority 1 expected impact: “eliminates 72% of negative sentiment.” The negative distribution is 45% / 36% / 19%.

No combination of the three categories yields 72%. The figure may refer to a different base (for example, BEV-attributable negatives), but the source does not say.

Reproduced as the report’s expected impact, flagged at the point of use.

387 qualified mentions, 54 undecided buyer responses, 35 direct competitor comparisons, 170 variant-specific discussions.

Whether the 54 responses are drawn from the 387 mentions, and whether the 170 variant discussions overlap either set, is not stated — so the four cannot be summed or treated as a funnel.

Reported as four independent bases throughout.

The same paragraph states that brand trust is 45% of negative sentiment and that it is “mentioned infrequently.”

Both statements cannot be read at face value. The likely reconciliation — low absolute volume but high decision impact — is implied by “irreversible decision-breaking factor”, but never stated.

Both reproduced; the tension recorded here rather than resolved by this site.

+28% conversion, “150% increase in 8 weeks”, “top 3 within 6 months” are stated without a baseline. No owning function or KPI is assigned to any of the four actions.

Without a baseline these read as targets invented at the point of writing; without owners they cannot be executed as published.

Every impact labelled “expected”, never presented as achieved; no owner column added.