"What's this worth" feels like a single question, but answering it well is actually four separate steps, done in order. Skip the first step and every step after it is answering the wrong question accurately.

1. Identify the exact thing first

A generic description prices a generic item, and the generic item is usually worth less — or more — than the specific one sitting in front of you. "A blender" and "a KitchenAid Pro Line 5-speed blender, red" are not the same lookup. Brand, model, and any variant that changes the price (size, material, edition, color in some categories, generation in electronics) all matter. Check for a model number, a tag, an engraving, or a stamp before you guess at what it is. The single most common pricing mistake is skipping this step: pricing "a watch" instead of the actual watch, and ending up with a number that has nothing to do with the item you're holding.

2. Find completed sales for that exact item

Once you know exactly what it is, look for what it has actually sold for — not what it's currently listed for. See sold price vs. asking price for why that distinction matters; the short version is that active listings run above what buyers actually pay, so a completed sale is the only number you can trust. Look for several recent sales, not just one: a single data point can't tell you whether it's typical.

3. Group by condition

A brand-new item and a well-worn one aren't the same market even if they're the identical model. Sort what you find into rough condition buckets — new, open box, refurbished, pre-owned, for parts — before you average anything. See item condition grades, explained for what each tier actually means. If most of the sales you found are for new items and yours is used, the number you're looking at doesn't apply to you yet — narrow it down to sales in roughly the condition your item is actually in.

4. Take the median, not the best case

Once you have several completed sales for the right item in the right condition, take the middle value, not the highest one. It's tempting to anchor on the best sale you found — it's the number you want to be true — but a single high outlier (a bidding war, an unusually motivated buyer) isn't what a typical buyer will pay for a typical one. The median of several real sales is a far more honest estimate than the best sale you can find with enough scrolling.

Where people go wrong

Almost every bad estimate traces back to one of three mistakes: pricing the wrong variant because the item wasn't identified precisely enough, treating one outlier sale as if it were typical, or reading an asking price as if it were a sold price. All three share the same root cause — skipping straight to a number instead of doing the identification work first. A generic search for "electric guitar" returns a price range wide enough to be useless; the same search for the specific make, model, and finish returns a tight, trustworthy one. The extra ten seconds spent identifying the exact item is what makes every later step actually mean something.

Every number in a good estimate should trace back to an actual listing you can point to — if it can't, it's a guess, not a price. That traceability is the whole point: a price you can check is worth more than a price you have to take on faith, whether you're the one buying or the one selling.