How to find out what something is worth on eBay using completed sales and real comparables

The most common mistake is using active listings as if they were sales. A product can be listed at 500 € for months without anyone buying it. That figure describes supply, not market acceptance.

The current Estudio de mercado (Product Research) page lets you look up data on millions of items sold over up to the last three years. It includes average sold price, sold price range, average shipping cost, formats, number of sellers and sell-through rate for recent searches.

Its FAQ explicitly distinguishes this tool from searching completed listings: sold completed listings cover the last 90 days, while Product Research can go back up to three years and shows the real price when a Best Offer was accepted. That difference stops you from reading the original visible price as the price paid.

In collectables, edition, year, manufacturer, variant and provenance can dominate the price. In electronics, capacity, battery, locks, accessories and physical condition can change it drastically. Matching must be category-specific.

The average price a tool offers can be useful, but an exceptionally high or low sale can distort it. With enough comparables, the median usually provides a robust reference; even so, you should look at the distribution and explain why you exclude an outlier.

Conceptual example: if six similar sales cluster between 90 and 110 € and one appears at 250 €, do not delete the 250 € automatically. Check whether it was a special edition, a larger lot or a different condition. If it was not, document that it is atypical.

A price from three years ago may be relevant for a stable antique and of little use for a phone that depreciates quickly. Weight recency according to the category.

Separate auction sales from fixed-price sales when the format has an effect. An auction with few bidders may close low; a fixed price may reflect a longer wait. If the tool offers a sell-through rate, use it as context on liquidity, not as a guarantee that your listing will sell.

Two sales at 100 € are not equivalent if one had free shipping and the other 25 €. Product Research shows shipping cost metrics; Barguvo must decide whether it analyses the item price, the total cost or both, and label them.

For the seller, the sold price is not the net amount received either: FEES calculates fees and other costs. Keep “gross sold” and “seller net” as separate fields.

If there is only one sale or the comparables are weak, widen the range, lower the confidence and explain what is missing. You can use auxiliary signals such as active listings or external estimates, but they must keep their label.

Barguvo should not show “47,83 €” when the evidence only allows it to say “approx. 35–60 €”. Visual precision cannot exceed the precision of the data.

Barguvo can help you organise the evidence, calculate scenarios and prepare a decision, but the final action belongs to eBay. It does not publish listings, bid, buy, pay or send offers or messages on your behalf.

When a figure, deadline or programme depends on the category, account, country or the listing itself, the data eBay shows in that specific flow takes precedence. This reference is dated 25 August 2026 and was designed to force dynamic data to be revalidated before a real listing.

Imagine you are looking up a console of a specific edition. Product Research returns twenty sales. First you remove lots with extra games, faulty units and other storage capacities. Then you separate new, refurbished and used. The set shrinks to eleven.

Of those eleven, three are from another market with very different shipping costs and currency. You can keep them as secondary evidence, but not give them the same weight. Another two have photos or titles that are not enough to confirm the edition. Six high-quality comparables remain.

Now you look at their dates. If five are from the last two months and one from two years ago, the old one can show historical context, but it should not dominate a category that depreciates quickly. You calculate the recent median, range and spread. If one sale sits very far away, you investigate before calling it an outlier.

Then you adjust for the unit you are valuing. If it includes the box and original controller and the comparables do not, the adjustment may be positive; if it has heavy wear or is missing an accessory, negative. Barguvo should explain the reason for the adjustment instead of adding a fixed universal percentage.

The result might be “comparable sales 185–215 €, median 199 €, medium-high confidence”. You can then derive a different listing range depending on the target speed, but that second figure must be labelled BARGUVO_ESTIMATE. Do not rewrite the observed sales to match your recommendation.

Finally, keep the time source. An analysis that is strong today can go stale. For technology, trend-driven fashion or volatile markets, refresh comparables before making a new decision. For very rare historical pieces, an old sale may still be valuable, but the uncertainty must remain visible.

To use this guide in a real transaction, first create a minimal valuation record. There is no need to copy all of eBay: keep only the data that changes the decision. To estimate a range, the core fields are completed sales, similarity, condition, recency, format and total cost. If one is missing, mark it as unknown instead of filling it with an assumption.

Every material claim should be traceable back to evidence: Product Research or a traceable comparable. If the data comes from the other party, label it as declared; if it appears directly in the listing or order, as observable; if you provide a photo or document, as supplied evidence. “Confirmed” should be reserved for what has actually passed an appropriate check.

Also distinguish the observation date from the date of the fact. A fee checked today may apply to a future sale under different conditions; a sale completed six months ago may still be comparable, but it is not a recent observation. This separation lets you update the guide without rewriting the history and avoids presenting old data as current.

When two sources conflict, do not automatically choose the figure that favours your goal. Prioritise the official source specific to the flow, check the country, category and date, and keep the conflict on record if you cannot resolve it. A specific checkout interface can be more specific than a general page; an old screenshot can be less reliable than the current help.

The result must express the level of certainty. If the data is solid, you can give a direct recommendation; if important elements are missing, widen the range or ask for a specific check. If the evidence exists but is not enough to decide, use “inconclusive”. That label does not mean the same as “unknown”: in one case the evidence is insufficient and in the other the data is missing.

Finally, do not turn Barguvo’s recommendation into the certified value. The assistant explains what evidence supports a decision, what uncertainty remains and what the next safe step is. The execution and the real contractual status belong to the eBay flow or, where applicable, to the parties and the applicable regulations.

This approach also makes SEO maintenance easier. If eBay changes a figure or a feature, the affected block and its review date are updated without altering stable principles such as separating listed price from sold price, reputation from authenticity, or platform protection from legal rights. The guide stays useful even if the interface changes.

Before closing the transaction, do a reverse reading: imagine that in a month you have to explain why you made that decision. If you cannot reconstruct it with the data you kept, traceability is missing. If you can do it with a few clear pieces of evidence, the process is documented well enough without turning an everyday purchase or sale into a disproportionate investigation.