Catawiki estimate: use it as a reference, not as a valuation or a guarantee

Barguvo distinguishes ASKING, FINAL_BID/SOLD, EXPERT_ESTIMATE, BARGUVO_ESTIMATE and PROFESSIONAL_APPRAISAL. ASKING is what someone is asking for; FINAL_BID/SOLD is a transaction result; Expert estimate is Catawiki’s range; Barguvo estimate is Barguvo’s own evidence-based calculation; appraisal is a professional valuation with its own scope and liability.

Two numbers can be identical and belong to different categories. A seller may ask 1.000 € and an auction may close at 1.000 €, but that does not turn the asking price into sold data. Label the type before using it.

Catawiki explains that its experts try to estimate the possible final bid value. To do this they may consider public and private information, including historical sales data, and attributes such as artist or manufacturer, origin, period, provenance, material, dimensions, rarity, type and condition.

The platform also points out that its experts do not carry out a general physical inspection and depend on the accuracy of part of the information provided by the seller. That is why the estimate has epistemological limits as well as market limits.

An auction depends on who takes part, when, how much they want the piece and what the competition is. The estimate may end up below or above the final price. Nor does it guarantee what a third party will pay if you resell the item.

Catawiki recommends consulting your own appraiser when market value is important and warns about the risk in purchases made for investment or speculation. This rules out any Barguvo copy along the lines of “guaranteed value” or “safe investment”.

Catawiki describes Gallery estimates for certain visual works and Retail estimates for some jewellery items and gemstones. These variants incorporate specific references —for example gallery ranges or comparable retail— in addition to the item’s attributes.

The appearance of a specific label does not change the fundamental rule: it is still an estimate with limits and not a guaranteed resale price. In addition, reports from external laboratories used in gemmology carry their own responsibility and Catawiki does not guarantee their accuracy.

For some items with Comprar ya, Catawiki states that it offers an estimated value based on demand and past sale prices of similar items. That confirms that sales history can feed internal models.

However, you should not turn that signal into an exhaustive public dataset. A closed page may show the final bid and serve as a verifiable comparable, but Barguvo will not promise a complete historical search tool unless authorised and verified access exists for that use.

For collectibles, compare identity, edition, period, provenance, condition, dimensions, rarity, quantity and date. For art add technique, authorship and size; for watches, reference, material, configuration and condition of components; for gems, report, treatment, weight and quality.

Give greater weight to recent and very similar comparables. If a sale differs on a material variable, do not discard it automatically: use it as a weak comparable or to understand the effect of that variable.

To build a useful sample, first define inclusion and exclusion criteria. For example: same reference, same material, size within a reasonable margin, comparable condition, sold in the last 24 months and non-exceptional provenance. If you change those filters after seeing the results to get the range you wanted, you introduce selection bias.

Date matters differently depending on the category. A mass-produced watch model may have a liquid market and rapid changes; a very rare archaeological piece may have few sales in a decade. “Recent” does not mean the same number of months for everything. Barguvo must show the age of the evidence and reduce confidence when the sample is scarce or heterogeneous.

You also need to normalise lots. A sale of ten coins cannot be compared directly with a single coin by dividing by ten if the grades differ. A set of books with one rare volume may concentrate almost all the value. Before calculating a unit price, check whether the items are interchangeable and whether the lot carries a premium or discount as a set.

When different currencies are involved, convert at the exchange rate for the relevant date or state explicitly that you are working in the original currency. A current conversion applied to an old sale can be useful for reading, but it should not hide the original figure. The same goes for inflation: adjusting may answer a different economic question from comparing observed nominal prices.

Final confidence must depend on coverage and agreement. Five high-similarity comparables forming a narrow range justify more precision than twenty vague results. If the evidence is poor, a wide range with low confidence is more honest than a central number with two decimal places.

An average can be dragged by a record or an abnormally low sale. The median is usually more robust, but it is no use either if you mix incompatible editions, sizes or conditions. Filter first; calculate afterwards.

Flag outliers and look for an explanation: exceptional provenance, important certificate, incomplete lot, poor description, restoration, market timing or few bidders. Do not remove a value just because it “gets in the way” of the range you expected.

Do not confuse spread with error. In markets for unique pieces it is normal for two similar sales to end at different prices because of timing, audience and presentation. A wide range can be an honest description of the market, not a failure of the model. Precision should increase only when the comparables genuinely converge.

If a Catawiki estimate and your comparables diverge widely, do not automatically pick the number you prefer. Investigate which attribute might explain the difference: provenance, rarity, condition, size, signature, certification, geographical market or simply an insufficient sample. The divergence is a signal to review the model, not to delete one of the sources.

When a closed Catawiki page shows the final bid, that figure can be labelled FINAL_BID/SOLD. It is stronger for valuation than an active asking price because it represents the observed close of a transaction, but on its own it does not prove a universal market price.

Record date, currency, category and comparable attributes. Do not reuse an old result without adjusting for context. And do not confuse the final bid with the buyer’s total cost, which adds Buyer Protection, shipping and possible import charges.

The official help states expressly that Catawiki does not provide a free valuation service. If you submit a higher-value lot and it is approved, the expert may provide an Expert estimate that is visible during the auction.

This avoids a tempting but incorrect SEO claim: “Free valuation on Catawiki”. If the user needs a valuation backed by professional liability for insurance, inheritance, litigation or tax purposes, they should look for the appropriate type of appraisal.

Also keep the date on which you checked each comparable and the state of the page. Markets change and an observation that is verifiable today may disappear or be updated; recording the snapshot makes it possible to know which evidence the Barguvo estimate actually used.