CarPriceLab / Methodology

The method
behind the meaning.

CPL-2.5: a transparent, comparable-based method for estimating vehicle market value from recent retail evidence. From vehicle selection to a traceable report.

SPECIFICATIONCOMPARABLESDISTRIBUTIONCONTEXT
01 / Define the comparison

A vehicle is more than a make and model.

CPL-2.5 matches make, model and registration year. Mileage is selected within ±30,000 km or ±30%, whichever is greater. Fuel and transmission match when supplied. Power is matched within ±8 kW or ±18%; known battery capacity within ±8 kWh or ±15%. The closer the match, the greater its contribution to the estimate.

02 / Select relevant evidence

Recency matters. So does provenance.

The source is AutoScout24. Every valuation uses prices actually observed within the past 90 days; a listing being seen again without a complete price does not refresh its price observation. Newer prices carry greater weight. Listings in EUR marked as VAT-deductible are normalized to netto using the source country’s standard VAT rate recorded by the importer, not a listing-specific rate. Conditional offers, unconfirmed VAT and incomplete price observations are excluded. Brutto adds your selected VAT rate to the same netto basis. Each report identifies its price basis, observation period and source listings.

03 / Limit extreme prices

Reduce the influence of the tails.

Repeated source IDs and probable relistings with the same seller and detailed vehicle fingerprint are grouped, keeping the newest observation. Within each country, prices beyond 4.5 median absolute deviations are excluded only when the remaining sample still has at least 10 observations. Prices are not trimmed across different country markets.

04 / Choose the valuation basis

Local when supported. Broader when needed.

Your target sale market is used when at least 10 comparable observations qualify. Otherwise the result explicitly identifies a European reference. A country needs at least five observations to contribute to that reference or appear in the country table. European calculations give each qualifying country equal aggregate weight.

05 / Describe the range

A distribution, not a promise.

The estimated market value is the weighted median; P25–P75 describes the central half of weighted comparable asking prices. It is a market range, not a statistical confidence interval or a forecast of the final sale price. Up to 60 matches per market are selected. Estimates are rounded to €100; source listing prices retain their observed values.

P25–P75 · retail referenceP50 / weighted median

Conceptual scale. P25–P75 is the typical market interval, not a price floor, ceiling or confidence interval.

06 / Keep the evidence

The evidence stays with the conclusion.

Every report carries a reference number, methodology version, selected market and observation dates. Its PDF preserves that assessment. No asking-price percentile is presented as a verified wholesale transaction price.

Evidence thresholds

Enough to show. Enough to use.

10

Overall valuation

Minimum valid comparables for an available result.

5

Country statistics

Minimum records to display a country median.

10

Primary country

Minimum records to use a country as the valuation basis.

90 days

Observation window

The same window for every valuation, based on the last complete price observation. More recent observations carry greater weight.

Methodology CPL-2.5. Match weight is 1 / (1 + 2 × distance), multiplied by exp(−days since observation / 90). Distance combines relative mileage difference and half the relative power difference; an unspecified comparison battery adds 0.25 when a target capacity is supplied.

CPL-W1 / Retail-based model

A modelled wholesale reference.

Derived from the lower retail segment

The same comparable netto prices are normalized to each country's retail midpoint. The weighted mean of the lower 10–30% segment is then applied to the selected retail value. This avoids treating a cheaper country's price level as a wholesale discount. No fixed dealer margin or cost deduction is applied.

Shown only with sufficient support

At least 20 effective comparables and three known sellers in the contributing segment are required. Concentrated samples and values that change by more than 5% when a seller is removed are suppressed. Retail valuation remains available independently. This is a retail-derived model, not an observed B2B transaction price; its range is not a transaction confidence interval.

From market evidence
to a commercial decision.

Listed price is not achieved price. The difference matters when a valuation informs a real commercial decision.

What the evidence can support

Relative positioning within a comparable listing sample, a view of dispersion, differences between represented markets and a transparent reference for further review.

Keep the commercial decision separate

The report estimates market value. Your purchase budget also depends on transport, preparation, selling and holding costs, as well as your margin target. Assess those separately from the market reference.

Additional analysis, on request

Discuss sale-speed research, transaction-price studies, market trends or equipment-focused analysis with us. Scope, data availability and delivery are agreed individually before an assignment begins.

Your next decision, better informed.

Bring context
to your workflow.

Tell us what you buy, value or build.
Let’s shape the right approach.

Start a conversation