The usefulHQ Used Price Index
What secondhand prices are actually doing, measured every day rather than guessed. We record the cheapest genuine listing for about 2,200 specific models every day, and this is the whole lot in one number, plus the same number for each category.
Last 7 days flat · last 28 days -0.2% · since 2026-08-07 -0.3%. Built from 97,857 day-to-day price pairs, of which 2,316 were discarded as changed listings rather than price moves.
What moved this week
Seven days to 12 October 2026, from the same chained index. 3 that made most of the week's move in a single day are left out, for the reason given under By category.
Fell most
- Sonos speakers -2.6%
- Strollers -1.8%
- KitchenAid mixers -1.6%
- Launch monitors -1.5%
- Edrum kits -1.5%
Rose most
- GoPros +2.0%
- Pizza ovens +1.8%
- Networking gear +1.8%
- Graphics cards +1.6%
- Galaxy Tabs +1.5%
By category
The furthest fall over the 9 weeks is Golf rangefinders, -12.6%. The biggest rise is Graphics cards, +8.5%. Categories with fewer than 8 tracked models are left out because a single listing would swing them. Those two, and the chart above, leave out the 21 categories marked mostly on, which made most of their move in a day or two: a category moves 2% in a day only about once in eighty days, and when it does it is usually the collector matching different listings rather than prices changing.
| Category | Models | 7 days | 28 days | Since 2026-08-07 |
|---|---|---|---|---|
| Golf rangefinders | 22 | -0.5% | -1.0% | -12.6% |
| Projectors mostly on 19 Aug and 3 Sep | 28 | -0.7% | -1.8% | -12.1% |
| Samsung Galaxy phones | 40 | -1.1% | -11.5% | -8.3% |
| Sonos speakers | 12 | -2.6% | +0.1% | -8.1% |
| Office chairs mostly on 16 Aug and 23 Aug | 10 | -2.3% | -3.4% | -7.5% |
| Impact wrenches mostly on 27 Aug and 15 Sep | 20 | +0.1% | -3.3% | -7.2% |
| Cameras | 44 | -0.3% | -3.2% | -6.6% |
| Drones | 19 | -1.4% | -6.4% | -6.5% |
| Wacom tablets | 19 | -1.4% | -5.3% | -6.5% |
| Acoustic guitars mostly on 12 Aug | 25 | +0.7% | +0.4% | -6.4% |
| Electric guitars | 40 | flat | -1.0% | -5.8% |
| Dyson vacuums mostly on 20 Sep and 1 Oct | 17 | -0.3% | -10.7% | -5.6% |
| Fishing reels | 29 | -0.1% | -4.5% | -5.1% |
| Camera flashes | 25 | -0.5% | -2.2% | -5.1% |
| Bass guitars | 26 | +1.1% | -1.3% | -4.8% |
| Gaming monitors | 16 | -0.8% | -1.7% | -4.2% |
| PA speakers mostly on 8 Aug and 5 Sep | 11 | -0.7% | -2.0% | -4.1% |
| Garmin watches mostly on 5 Sep and 19 Sep | 11 | -3.5% | -1.9% | -4.1% |
| Security cameras mostly on 3 Sep and 25 Sep | 21 | +1.3% | +7.1% | -3.6% |
| Embroidery machines mostly on 29 Sep | 15 | +0.2% | -2.7% | -3.6% |
| Blenders | 22 | flat | -1.5% | -3.4% |
| Welders | 16 | flat | -0.4% | -3.2% |
| Digital pianos | 19 | -0.2% | -2.5% | -2.9% |
| Guitar pedals | 40 | -0.1% | -1.4% | -2.9% |
| Golf clubs mostly on 3 Sep | 25 | -1.2% | -2.5% | -2.8% |
| Galaxy Tabs mostly on 19 Sep | 16 | +1.5% | -1.3% | -2.7% |
| CPUs | 41 | -0.1% | -1.4% | -2.5% |
| Coffee grinders | 23 | flat | -3.4% | -2.5% |
| GoPros | 18 | +2.0% | -4.6% | -2.2% |
| Test equipment | 25 | -0.1% | -0.4% | -2.1% |
| DJ controllers | 27 | flat | -0.9% | -2.0% |
| AV receivers | 19 | -4.0% | -2.7% | -2.0% |
| Gimbals | 16 | -0.3% | -0.8% | -2.0% |
| VR headsets | 10 | -1.3% | -1.9% | -1.9% |
| Soundbars | 17 | -0.8% | -1.5% | -1.9% |
| NAS drives | 24 | -0.1% | -0.5% | -1.8% |
| Strollers | 10 | -1.8% | -0.6% | -1.6% |
| AirPods | 8 | flat | +0.7% | -1.6% |
| Ham radios | 22 | -0.2% | -1.4% | -1.6% |
| Metal detectors | 23 | -0.8% | -1.2% | -1.5% |
| Launch monitors | 17 | -1.5% | +1.1% | -1.5% |
| Saxophones | 19 | +0.3% | -1.1% | -1.3% |
| Designer sunglasses | 25 | flat | -1.5% | -1.3% |
| Camera lenses | 42 | +0.7% | +1.5% | -1.1% |
| Audio interfaces | 25 | -0.3% | -0.9% | -1.1% |
| Rowing machines | 17 | flat | -0.5% | -1.0% |
| Synthesizers | 40 | flat | -0.6% | -0.9% |
| Watches | 29 | +0.9% | -0.2% | -0.9% |
| 3D printers | 28 | -0.1% | -0.9% | -0.8% |
| Chainsaws | 22 | -0.6% | -0.6% | -0.7% |
| Sewing machines | 24 | flat | +1.1% | -0.6% |
| Air compressors | 16 | flat | flat | -0.5% |
| Business laptops | 42 | -0.7% | -2.3% | -0.3% |
| Designer handbags | 40 | -0.3% | +0.2% | -0.3% |
| Studio monitors | 41 | flat | flat | -0.2% |
| Headphones | 29 | -0.8% | -0.5% | -0.2% |
| Printers | 19 | +0.3% | +1.0% | +0.1% |
| Espresso machines | 29 | -0.2% | +0.1% | +0.1% |
| Sneakers | 41 | +0.1% | +0.2% | +0.3% |
| Robot vacuums | 29 | flat | -1.1% | +0.3% |
| Telescopes | 25 | flat | flat | +0.3% |
| Apple Watches | 11 | flat | +0.5% | +0.5% |
| Film cameras | 42 | +0.8% | +0.9% | +0.5% |
| iPads | 40 | +0.2% | +2.0% | +0.6% |
| Exercise bikes | 18 | +0.4% | -0.3% | +0.6% |
| Treadmills | 34 | flat | +0.6% | +0.6% |
| Turntables | 27 | -0.4% | flat | +0.9% |
| Guitar amps | 40 | -0.2% | -0.1% | +0.9% |
| Bench saws | 19 | +0.7% | -0.3% | +0.9% |
| Pizza ovens | 12 | +1.8% | +1.0% | +1.0% |
| Portable power stations | 25 | flat | +0.4% | +1.2% |
| Binoculars | 22 | +0.5% | +1.9% | +1.3% |
| Gaming laptops | 11 | +0.5% | -1.9% | +1.3% |
| Fitbits | 19 | +1.0% | +1.7% | +1.7% |
| Microphones mostly on 23 Sep | 41 | +0.8% | +2.9% | +2.0% |
| Electric scooters mostly on 12 Aug and 15 Sep | 11 | flat | -0.1% | +2.1% |
| iPhones mostly on 3 Sep and 16 Sep | 40 | flat | +8.8% | +2.4% |
| MacBooks | 40 | -1.1% | -5.0% | +2.5% |
| Generators mostly on 23 Sep | 17 | flat | +2.8% | +2.6% |
| Edrum kits mostly on 12 Aug | 19 | -1.5% | -0.7% | +2.6% |
| Networking gear | 24 | +1.8% | +2.5% | +3.2% |
| Mechanical keyboards mostly on 24 Sep | 24 | -0.4% | +3.0% | +3.6% |
| KitchenAid mixers mostly on 15 Sep and 23 Sep | 8 | -1.6% | +5.2% | +5.2% |
| Cordless drills mostly on 12 Aug and 23 Aug | 24 | -0.2% | +1.9% | +5.2% |
| Game consoles | 22 | flat | +2.9% | +5.9% |
| LEGO sets mostly on 14 Aug and 31 Aug | 44 | -0.1% | -0.6% | +6.6% |
| Graphics cards | 58 | +1.6% | +2.8% | +8.5% |
| Google Pixel phones mostly on 19 Sep and 24 Sep | 22 | -0.4% | +10.1% | +10.6% |
Download the whole index as CSV, one row per day per category, updated with this page. The same file, with a commit for every day it has ever changed, is at GitHub, so any figure quoted from this page can be checked against the day it was published. A day's figures are not recomputed once published: when a listing is later found to be wrong, the correction counts from that day on, and any change to the method is announced here as a revision.
Monthly reports
Each month, fixed once it ends: September 2026, August 2026.
How it is built
The index is chained, which means it measures each day against the day before rather than against a fixed starting month. Every model that has a price on both days contributes one ratio, and the index moves by the trimmed geometric mean of those ratios, dropping the top and bottom 10% so one silly listing cannot swing a category.
A day-to-day move on one model of more than 25% up, or 20% down, is thrown away rather than counted. Those are the same size of move: a 20% fall is exactly undone by a 25% rise. That size of jump is almost never a price change. It is the collector matching a different listing, and the reason this rule exists is worth stating plainly: a first attempt at this index, chained from a fixed August baseline, showed acoustic guitars down 32%. They had not fallen at all. Our own filters improved on 12 August and stopped pricing guitars from listings that were never guitars, so the "fall" was us getting better at our job. An index that cannot tell its own corrections from the market is worse than no index.
Prices are the cheapest Buy It Now listing on eBay US for each model, in US dollars, excluding parts, accessories, broken units and bundles as far as our filters can tell. No survey, no dealer submissions, no estimates. If a model has no price on a given day it simply does not vote that day.
Revised 30 September 2026. Until then the rule was 25% either way, which kept a 24% fall but threw away the 33% rise that reversed it, so a price bouncing between two listings was pushed down one bounce at a time. The day's ratios were also averaged arithmetically, which pushes the other way when prices bounce. Both were fixed together and the whole history was recomputed. The whole market barely changed (-0.3% to -0.2% on 26 September) but a few categories did: Samsung Galaxy phones had read -24%, and read about -5% under the fixed method.
What it does not tell you
It is a cheapest-listing index, not an average of sale prices, so it tracks where the floor of the market sits rather than what a typical buyer pays. It covers consumer goods that people resell online, weighted equally by model rather than by how many change hands, so a category with four models counts as much as one with fifty. And it starts in August 2026, which is a short history. Treat a one-month move as news and a six-week move as a hint, not a trend.
Where it sits next to the CPI
The US Consumer Price Index has long priced used cars and trucks, and until recently little else bought secondhand. Secondhand clothing joined it in early 2025, and the Bureau of Labor Statistics says it is researching others, naming electronics, books and furniture (Monthly Labor Review, May 2026). This index is no substitute for that work: it prices the cheapest listing rather than what people pay, and it weights models equally rather than by what households spend. What it offers is a daily reading of the goods that research is looking at, most of them electronics, while the official series does not yet cover them.
If you are writing about this
Everything here is free to use. What is probably more useful than the headline number: we hold a daily price for about 2,100 specific models across 94 categories, so most questions of the shape "has X got cheaper" can be answered with a real figure rather than an estimate, usually the same day you ask.
Things we are asked for and can pull: one category over any window, one model's full history, the biggest movers in a period, a like-for-like comparison of product generations rather than a median across ages, or the raw rows behind any figure so your desk can check it. If a number here does not hold up, say so and we will correct the page.
What we will not do: dress a small sample as a trend, or give a figure without saying what it is measured against. Several categories are deliberately left out of the table because eight models is not enough to say anything.
Related
How much each category loses secondhand, comparing used prices with what the same models cost new. What is it worth, a lookup for one model at a time. Price drops, the models that moved most this week.