How to Price Resale Items With Sold Comps and Market Data

A fast sale can mean strong demand, an excellent listing, or a price that was too low. You cannot tell which from speed alone. A defensible list price starts with comparable sold items, then adjusts for condition, completeness, platform fees, shipping, and your minimum acceptable profit.

This is the compact workflow. If you manage a large catalog and need scheduled repricing across many SKUs, use the AI dynamic pricing workflow. This page is for one-item comp selection and price-floor math.

The Five-Minute Pricing Record

Before listing, save:

• exact item identity: brand, model/style, variant, size, color, material, year/era when verified;
• condition and every defect;
• accessories, packaging, paperwork, or missing parts;
• five to ten close sold results with URL, sold date, price, shipping, and condition;
• excluded results and the reason for each exclusion;
• marketplace fee basis and packed shipping cost;
• list price, offer floor, and next review date.

Use eBay Product Research when available. eBay says it can show actual accepted-offer prices, sold ranges, shipping cost, and up to three years of sales data; public completed-listing results cover a shorter window and can hide the accepted offer. The sold-listing link generator is a faster starting point, but the filters and exclusions are still your responsibility.

Step 1: Define the Item Before Searching

Bad searches create bad comps. Write a one-line identity:

Brand + exact model/style + variant + size/spec + condition + included pieces

Examples:

• Starter Chicago Bulls pullover jacket red black men's L used light cuff wear
• Sony WM-D6C cassette recorder tested play record no case
• Levi's 501 made in USA button 552 measured 32x30 faded no holes

Do not broaden the search until you know which detail is preventing matches. A different model, licensed graphic, voltage, production run, size, or missing accessory can be a different market.

Step 2: Collect the Raw Sold Set

Save every plausible result before choosing favorites. For each result record:

An asking price is not a comp. An active listing tells you about competition, not what a buyer paid.

Step 3: Exclude Mismatches With a Written Reason

Exclude a result only when you can name the mismatch:

• wrong model, style, variant, size band, voltage, edition, or production run;
• materially better or worse condition;
• new-with-tags versus used;
• bundle/lot versus one item;
• complete-in-box versus missing parts;
• refurbished/serviced versus untested;
• local-pickup result versus shippable national sale;
• suspicious duplicate, cancelled result, or unclear identity.

Do not remove a low sale merely because it lowers the median. Check whether its condition, shipping, title, photos, or auction format explains it. If not, it is market evidence.

Worked Example: Vintage Starter Bulls Pullover

This is an illustrative calculation showing the method. The dollar figures are not current market claims; replace them with sold results collected on your pricing date.

Item: 1990s-style Starter Chicago Bulls pullover, men’s tagged L, measured chest and length recorded, used, light cuff wear, no holes, no original tags.

Raw set

Kept sold prices are $52, $58, $65, $68, $72. The median is $65. Do not stop there: normalize shipping and condition.

Normalize the results

• Results C and F included seller-paid shipping. Compare seller net, not headline price.
• Results D and F were cleaner than the subject item. A condition adjustment should move the subject below those outcomes.
• B and E are the closest condition/size matches and deserve the most weight.

Assume your current marketplace fee calculator shows $F in fees at the intended price, packed shipping is $S, packaging is $P, acquisition cost is $C, and the return/uncertainty reserve is $R.

net proceeds = accepted sale price - F - S - P
profit = net proceeds - C - R

Use the flip profit calculator with the actual marketplace and shipping quote. The article’s sample sales should never replace a fresh search.

Step 4: Calculate the Price Floor

Your floor is not “the lowest comp.” It is the minimum accepted price that still meets your business rule.

offer floor = acquisition cost
            + prep/repair cost
            + packaging
            + seller-paid shipping
            + marketplace fees at the offer price
            + return/uncertainty reserve
            + required profit

Because percentage fees depend on the accepted price, use the marketplace-specific calculator rather than subtracting a remembered rate. For current platform comparisons, use the platform fee comparison tool.

If the calculated floor sits above the close sold range, you overpaid, underestimated fulfillment, or chose the wrong platform. Raising the list price does not change that evidence.

Step 5: Choose List Price and Offer Room

Set list price from three facts:

1. the normalized close-comp range;
2. your condition and listing evidence relative to those comps;
3. the offer floor.

A list price above the median can be reasonable when the item is better, complete, scarce, or supported by superior evidence. It is not reasonable because you “want room.” Offer room has to stay inside the observed buyer range.

Write the floor before offers arrive. That prevents a live negotiation from rewriting your economics.

Step 6: Use a Markdown Rule, Not Panic

Set review dates when you publish the listing:

Seasonal items need a calendar-aware rule. A coat listed in spring should not be judged against peak autumn velocity, but it still needs a storage-cost decision.

Step 7: Separate Price Problems From Listing Problems

Before cutting price, check:

• correct category and item specifics;
• exact model/style/size in the title;
• clear first photo and complete measurements;
• condition and flaws disclosed;
• competitive delivery cost and handling time;
• enough demand in current sold data;
• buyer-facing fees or shipping that make checkout expensive;
• whether the item belongs on a different platform.

The listing title optimizer helps with title structure. The competitive price monitoring guide is the next step when multiple direct substitutes are moving around your listing.

Platform Differences Without Guessing

The same item can produce different gross prices and different net outcomes. For each platform, compare:

• close sold count and date;
• accepted sold price;
• seller fee and payment processing;
• buyer-facing fee and shipping;
• seller-paid label or free-shipping cost;
• return rules and category risk;
• audience fit and listing labor.

Do not assume eBay always pays most, Poshmark always pays more for fashion, or Mercari always sells faster. Those are hypotheses to test with the exact item.

What Not to Do

• Do not average active asking prices.
• Do not use one exceptional auction as the market.
• Do not mix restored, tested, untested, and parts units.
• Do not apply a universal percentage premium for better photos or a familiar tag.
• Do not hide shipping inside “free shipping” when calculating net.
• Do not lower below the written floor because a buyer says another listing is cheaper.
• Do not keep repricing a listing whose real problem is identification or demand.

Sources and Freshness

• eBay Product Research, checked July 21, 2026: sold-price range, actual accepted-offer prices, shipping cost, filters, and up to three years of data.
• Platform fees and carrier rates are dynamic. Use the current fee page, sold receipt, and packed shipping quote for every calculation.
• Worked-dollar figures in this guide are explicitly illustrative. They demonstrate selection, exclusion, normalization, floor, and markdown logic; they are not a live appraisal.

Frequently Asked Questions

How many sold comps do I need?

Five close matches are a useful minimum when the market has them; ten is better for common inventory. For rare items, use fewer only with a larger uncertainty reserve and more specialist evidence.

Should I include auction results?

Keep them visible but label the format. A poorly timed auction can close below a fixed-price market, while a bidding contest can create a high outlier. Compare like with like.

What if the accepted offer is hidden?

Use a research tool that exposes the actual price when available. If you cannot see it, mark the result as price unknown instead of treating the crossed-out ask as the sale price.

When should I lower the price?

After you refresh sold data and rule out identification, category, photo, condition, shipping, and audience problems. Lower only to a price that stays above the documented offer floor, unless you consciously choose liquidation.

When should I use dynamic pricing instead?

Use dynamic pricing when a large catalog needs repeatable review triggers. Keep this exact-comp workflow as the underlying evidence; automation should not turn weak comps into confident prices.