US specialty jewelers add 5.7% in September on fewer units
Tenoris recorded close to two years of uninterrupted monthly growth, with year-to-date revenue up 8.5%. Average spend per item rose 11%; sales of lower-priced jewellery fell 11% and natural diamond jewellery slipped 0.6%.
By The Retail & Technology Desk
What changed
Natural diamond jewellery turned negative. Revenue at US specialty jewellers rose 5.7% in September, but finished natural-diamond jewellery sales fell 0.6% and lower-priced jewellery fell 11%.
What it means
- If you set assortment for Q4: The revenue sits in the upper price bands and the units are leaving the lower ones. A shelf built for last year's unit mix will under-earn.
- If you sell natural against lab: Natural jewellery revenue went negative in the same month lab-grown unit demand rose about 26%.
- If you plan cash: Average spend per item rose 11%. Fewer transactions carrying more money each is a thinner, more fragile top line.
- What this is not: A channel-wide read. Tenoris measures US specialty jewellers, not chains, department stores or marketplaces, and 0.6% is inside one month's noise.
Key figures
| US specialty jeweller revenueSeptember, year on year | +5.7% |
|---|---|
| Average spend per itemSeptember | +11% |
| Lower-priced jewellery salesSeptember | −11% |
| Finished natural-diamond jewelleryends an extended streak | −0.6% |
Source: Tenoris monthly data for September 2026, reported by analyst Edahn Golan on 7 October 2026 and published by JCK the same day
Revenue up, units down
Two years running. Golan put September at 5.7% above a year earlier and the streak at close to two years of uninterrupted monthly growth. Year to date, revenue is up 8.5%.
The ladder tilted. Average spend per item rose 11% while unit demand weakened. Lower-priced jewellery sales fell 11% at the same stores that were growing revenue.
| Line | September change | Note |
|---|---|---|
| All specialty jeweller revenue | +5.7% | Year to date +8.5% |
| Average spend per item | +11% | Unit demand still falling |
| Lower-priced jewellery | −11% | Revenue rose anyway |
| Finished natural-diamond jewellery | −0.6% | Ends an extended streak |
| Natural-diamond jewellery, incl. loose | −1% | Spend per stone +11% |
| Lab-grown diamond jewellery demand | +26% approx. | Loose lab-grown revenue down a 5th month |
Every row is Tenoris's own September figure as JCK prints it. The approximation on lab-grown demand is the source's own word, 'nearly 26%', and is carried as approximate rather than rounded to a false precision. No row is Carat Capital's arithmetic.
Natural diamonds slipped
Minus zero point six. Finished natural-diamond jewellery sales fell 0.6% in September, ending an extended growth streak. Including loose stones, the natural-diamond total fell 1%.
Buyers spent more. Average spending per stone rose 11% among the consumers who did buy. The money per transaction held while the number of transactions did not.
Lab-grown splits two ways
Demand up, revenue down. Lab-grown diamond jewellery demand rose close to 26% year on year. Revenue from loose lab-grown stones fell for a fifth consecutive month as prices kept dropping.
One quote, attributed. Golan described a market “increasingly divided by price point”, which is the clearest short statement of what the month's two directions mean together.
Luxury read points down
Citi says minus six. JCK cites Reuters of 6 October for a Citi credit-card analysis. It puts US luxury purchases down 6% in September, after 4% falls in July and August.
Jewellery held better. Citi found watches and luxury jewellery weakening further inside that figure, while jewellery-store data stayed positive. The two instruments measure different buyers.
What to watch
- Early November 2026Tenoris's October reading. A second negative month on natural-diamond jewellery would end the streak properly rather than nicking it.
- Through Q4 2026Whether lower-priced unit demand stabilises. Eleven per cent down is the number that decides entry-level assortment for 2027.
- Mid-October 2026Citi's next luxury card read, which is the only outside check on whether jewellery keeps outperforming the wider category.
The story so far
Go deeper
What would change this call
An October reading that puts unit demand flat or positive while revenue holds. The whole reading here rests on revenue and units moving in opposite directions; if they re-converge, this is a one-month mix effect and not a bifurcation.
How the Tenoris series is built
Tenoris aggregates point-of-sale data from US specialty jewellers and publishes monthly year-on-year changes in revenue, unit demand and average spend, broken out by category and price band. It is a measure of sell-through at independent and specialty doors, not of the whole US jewellery market: chains, department stores and online marketplaces sit outside it. Comparisons are to the same month a year earlier rather than to the prior month, so a figure can be positive against a weak base while trading is flat in sequence.
Method
JCK's article was fetched to raw bytes, HTTP 200, 108,248 bytes, saved to newsroom/sources/2026-10-08/jck-tenoris.html, and the figures were read out of the article's own body paragraphs rather than from a whole-page grep, because the page carries a popular-stories rail and a supplier-news block whose figures belong to other stories. Exact-digit greps on that saved file: 5.7% two hits, 8.5% one, 11% three, 0.6% one, 26% one, 6% three. The quote attributed to Edahn Golan was counted at seven words before filing. The Citi figure is attributed as JCK attributes it, to Reuters of 6 October, and was not re-opened at Citi or at Reuters, which is stated in the article rather than smoothed. Archive check, searched in website/content/articles.json by entity: 'Tenoris' returns five prior pieces, each opened and read -- more-dollars-fewer-boxes and pearls-print-a-quiet-win of 21 July, average-ticket-carries-the-half of 26 July, china-lowest-since-2004 of 3 August and the-fifteen-hundred-line of 8 August. All five cover the first half of 2026 or June 2026. This item covers September 2026 monthly data. SAME ENTITY, DIFFERENT PERIOD -- ADVANCED, not a repeat: ours covered H1 2026 and June 2026, this covers September 2026.