Choose sold-price evidence for one sealed resale unit

1. Define the unit before comparing prices

The fictional target is Orbit Series One, retail edition, English, factory sealed, one display containing 12 packs. Its internal SKU is ORB-S1-EN-12. A single pack, opened display, collector edition, or six-display case is a different comparison group.

Write down the maker or product identifier, edition, language, condition, package contents, selling channel, currency, and unit you mean. Your internal SKU helps organize your own records; a seller’s SKU or similar title alone does not prove an exact product match.

For each candidate, retain its real source reference, actual sale date, date you checked it, item amount, buyer-paid shipping, and any offer notation. Check the description and condition rather than relying on the headline or a broad average.

2. Keep, separate, or investigate each candidate

Assume the invented records use the same channel and USD. “Exact target” means the same edition, language, sealed condition, and single-display format. Row labels are teaching references, not marketplace transaction IDs or links.

Reasons matter more than whether a number is convenient. Row F is outside the sealed group because it is opened, not because its price is low. Row G has a product mismatch, not merely a high price. Do not remove a genuine close match just to improve the result.

Dividing row E’s $804 item-plus-shipping total by six gives $134 per display mathematically. It does not establish what one display would sell for on its own. Keep the case format separate unless you have a supported reason to compare that buying context.

3. Distinguish a displayed ask from an actual accepted price

An accepted-offer label does not establish the amount paid. eBay’s Offers to Buyers and Best Offer FAQ says the original Buy It Now amount can remain in the search-history view described by that FAQ after a discounted purchase. Its Product research guidance describes actual sold-price information including accepted Best Offers. These are different views; check the amount and record you are using.

In invented row C, the actual accepted item amount is explicitly given as $148, so use that amount. In D, only the original $175 ask is known. Leave D’s actual item amount unknown until it can be established; do not infer an offer discount or treat the ask as a completed-sale price.

For a real record, retain how and when you checked the amount. A reference to research is not proof that the particular transaction, condition, or unit matched your target.

4. Choose one comparison basis before doing the arithmetic

Keep the actual item amount separate from buyer-paid shipping. For a known single-unit record, their sum is a pretax buyer amount. It excludes sales tax collected from the buyer and is not the seller’s net payout.

Do not blend A’s $142 item amount with C’s $154 item-plus-shipping amount in one summary. The same label must mean the same components for every included row. Unknown shipping in I is not $0, so its comparable buyer total remains unknown.

For a planned free-shipping sale, your item price is also the buyer’s pretax amount, but your label cost, fees, and other costs still affect what you keep. eBay’s selling-fee guidance also describes fee bases that can include shipping and tax, with exceptions. The worksheet’s pretax amount is not automatically the fee base or net proceeds.

5. Describe the sample without turning it into a forecast

Use the same fully known A/B/C set for this paired comparison. Row J is also fully priced but is kept outside this set as separate historical context because it sold in July 2024. That is the example’s stated choice, not a universal recency cutoff. Sorted item amounts are $142, $148, $149: a sample median of $148 and range of $142–$149. Sorted pretax buyer amounts are $149, $150, $154: a sample median of $150 and range of $149–$154.

The two summaries describe different components of the same three invented transactions. Neither tells you what your unit will sell for. A sample median is not an automatic expected-price input; the range is not a guaranteed floor, statistical confidence interval, percentile, or sale probability.

Row I has a known item amount, so it can inform a separately labeled item-only review. It stays outside this paired set because shipping is unresolved. If you change which rows enter a summary, record that set and its count rather than comparing different groups silently. D cannot enter an actual-price summary while its accepted amount is unknown.

Three rows are not a required or sufficient sample size. If close records are scarce, say so. Preserve unresolved candidates and the reason they are unresolved. Do not invent price assumptions to make a sparse worksheet look complete. A lower-price scenario needs its own reasoning; it does not automatically equal the smallest recorded sale.

6. Keep the sale date and date checked separate

Each candidate needs its transaction date as well as the date you reviewed it. The worksheet’s assumed check date is October 30, 2026; row J still describes an invented July 2024 sale. Checking it on that assumed date does not make the underlying sale recent.

Keep older records separately when their relevance needs explanation. There is no universal time window that makes every product comparable. Note changes in edition, supply, condition, channel, and the available evidence rather than automatically treating a dated sale as current.

ResaleCraft’s Date checked records when you checked the evidence. It does not hold a separate sale date for each comparable or evaluate all the underlying transactions for you. Keep that fuller history in your own worksheet or research record.

7. Record only what the current pilot supports

The opportunity form offers three evidence inputs:

• Price-evidence source: a short source description or reference. In manual entry this is limited to 500 characters; it is not a structured table of comparable sales.
• Date checked: when you actually checked the evidence, distinct from each transaction’s sale date.
• Evidence confidence: your High, Medium, or Low assessment of evidence quality. The labels describe direct recent comparisons, relevant but imperfect evidence, or thin indirect evidence; they are not statistical confidence levels or an automatic reliability score.

For the invented worksheet, a source summary could identify the target SKU and worksheet reference, state that item and shipping amounts are separate, and preserve that offer or shipping amounts need follow-up. For a real decision, substitute a genuine reference and honest checked date. Keep the richer worksheet yourself; the current form and opportunity CSV do not provide an evidence-notes field, attachment upload, or structured-comparables import.

The CSV fields are evidence_source, evidence_observed_at, and confidence. Use the CSV preparation guide for their formats. A valid draft or import can still lack the facts needed for a buying decision.

If the required source or expected/lower sale price is unsupported, keep it unknown; those missing inputs leave the recommendation as Needs information. A missing checked date, evidence older than the configured age rule, or Low confidence instead produces warnings; when the hard buying rules pass, those warnings can lead to Wait. A confirmed hard-rule failure takes Skip precedence. Invalid or future checked dates need correction, not a confidence upgrade.

The app uses your sale-price assumptions and channel-cost settings. It has no separate buyer-paid-shipping input in these evidence fields and does not automatically turn a worksheet median into a price. Choose and document the basis for your estimate, and keep selling-cost assumptions consistent with it. See the maximum-buy-price example for the next calculation and the buying checklist for the other facts a decision needs.

Sources and scope

Primary references checked October 30, 2026: eBay: Product research; eBay: Offers to Buyers and Best Offer; and eBay: Selling fees. Check the current source and the specific record you can access when doing real research.

The worksheet, dates, figures, and SKU are invented. This guide does not authenticate sealed products, value inventory, predict a sale, or establish a statistical model. The invitation-only ResaleCraft pilot uses manual entry and CSV workflows in USD; it does not fetch comparable-sale records or connect to marketplace accounts. Entering a source reference cannot authenticate the product, grade its condition, or verify that an outside sale occurred.