What a crystal's price is really made of

A crowded market with a very long tail

Of 28,309 shops that show up in crystal searches, only about 2,600 are clearly crystal specialists. The top 1% of shops hold 36% of all listings.

If you open a crystal shop, you won't be first. We searched Shopify's product catalogue the way a buyer would, across the US, Canada, Australia and the EU. 28,309 shops came back. That is a sample of what a buyer can find, not a census, so the true number is higher.

Most of those shops aren't crystal shops. Jewellers, gift shops, candle makers and home-decor stores carry a few stones. Only about 1 in 12 is clearly a crystal specialist, but those specialists hold 44% of what a buyer sees.

The market is lopsided. The median shop showed us 2 listings; the average showed 11. Among specialists the middle shop lists 17 products and the average 55. A small group of large catalogues sets what most buyers see first.

Crowded doesn't mean cheap

You'd expect the stones everyone sells to be the cheapest. Across 661 stone-and-shape combinations, how many shops carry something told us almost nothing about its price. The exceptions are stones few shops carry at all: moldavite sits in about 1 in 6 specialist shops and asks about 2.5 times the usual price for its shape. Charoite, aquamarine and topaz are in fewer than 15% of shops.

Where you sell changes the number on the tag

Canadian and Australian shops look about 25% cheaper than US shops once you convert to US dollars. In their own currency the price tags are about the same. For stones, the gap is the exchange rate. Jewellery is different: outside the US it really does ask about half.

Shop names carry a signal too. Shops with "witch" or "moon" in the name ask 15 to 20% below the market for the same stone and shape. "Crystal" in the name makes no difference.

There is a going rate, but it's a range, not a number

For the same stone, form and size, the shop at the 75th percentile asks about 3× the shop at the 25th. Match the pieces more closely and the gap shrinks to about 2.5×.

The first thing anyone pricing crystals notices is that nobody agrees. That's partly true and partly an illusion. Line up every listing called "amethyst sphere" and the spread is enormous, because a 2 cm marble and a 20 cm display piece share a name. Narrow it to the same stone, the same form and the same size, and the middle half of shops sit within about 2.5 to 3 times of each other.

Big pieces cost less per gram

A piece under 50 g asks a median of about $1 a gram. A piece over 2.5 kg asks about 9 cents. Shape matters too. A sphere's price grows faster than its size: double the diameter and the price roughly triples. Double a tower's height and the price roughly doubles. Tumbled stones barely move. So price large pieces by the piece, not by the gram.

Prices end in round numbers. In the US, 68% of prices are whole dollars and only 14% end in .99. The most common US prices are $20, $25, $45 and $30. Exactly $100 is three times as common as its neighbours, and almost nobody prices between $100.01 and $100.99. Europe is different: only 46% whole numbers, and .95 is the charm ending.

Mostly about who is selling it

Knowing the stone explains 8.5% of the price. Adding form and size takes it to 30%. Knowing which shop is selling it explains 39% on its own.

This is the core test of the report. We took 93,964 listings of stones, spheres, towers and clusters from 1,114 shops that each had at least 20 in our data. Then we added one fact at a time and measured how much of the price each one explains.

The order doesn't change the answer. Averaged over every possible order, the shop is half of everything we can explain, the form a fifth and the stone 8%. It held on listings the model hadn't seen, on listings that state their size, and on mass-produced shapes like spheres and towers.

How far apart are shops?

For the same stone, form and size, a shop at the 90th percentile asks 2.7 times the market. A shop at the 10th asks 0.39 times. That's about a sevenfold gap. Even the 75th and 25th percentiles are 2.8 times apart.

A shop's price level is a position, not a mood. Split a shop's listings into two random halves and they give almost the same answer. A shop that is dear on amethyst is dear on fluorite and selenite.

What the expensive shops share

Less than you'd think. The top tenth of shops name an origin more often (43% of listings vs 18%), make fewer healing claims (39% vs 71%), carry fewer stones and sell more mineral specimens. They don't write longer descriptions and they don't disclose treatments more often. All the listing habits we measured predict only about a sixth of which shops are expensive. Most of a shop's price level is something you can't see in one listing.

One honest limit: "which shop" also absorbs real quality differences that listings don't state (84% give no size in the title). So the 7× is an upper bound on pure pricing habit. Chapter 07 shows the other side: when shops sell the identical product, they agree much more closely.

Trust and audience

Not in anything we could measure. Shops with reviews, a review app or a big social footprint price no higher than shops without. Everything visible explains about 22% of a shop's price level. The other 78% is something we can't see from outside.

Chapter 03 found that the shop sets the price. The natural next guess is that shops with more trust and a bigger audience charge more. We went back to 3,000 shops and read what a buyer would see: whether product pages run a review app and how many reviews they show, which social networks the shop links, whether it promises free shipping or a guarantee, and how many products the whole store really carries.

A review app, Instagram, TikTok, Facebook, Pinterest, three or more networks: none of them moved the median shop's price level by more than a few points. Products that display a review count ask the same as those that don't (+0.1%). The exceptions are small and worth a second look: shops linking a YouTube channel sit at the market (1.00×) against 0.79× for the rest, and shops that state a guarantee or show trust badges sit at about 1.02×.

Bigger catalogues don't price lower

You might expect huge stores to win on volume and price low. They don't. Shops with 10,000+ products price above the market; mid-sized stores of a few hundred products sit lowest. Many of the giant catalogues are mineral and gem dealers, which says more about what they sell than how many.

So what is the 78%? We can't prove it from public pages. The likely answer is the things a crawler can't count: the size and loyalty of a shop's real audience, what a buyer remembers about it, the quality of what it actually ships, and whether it sells collector specimens or gifts. Having a review app is easy. Being the shop people trust is not, and it doesn't show up as a checkbox.

Limits: review counts were readable on only 8.5% of product pages, because most review apps load them with JavaScript. We saw whether a shop links a social network, not how many followers it has. And we only have asking prices, so a trusted shop that sells more at the same price would look identical here. This chapter can't show trust doesn't matter. It shows the visible badges of trust aren't what sets the price.

Three pairs, three different answers

Across the three, more reviews line up with a higher price once and a lower price twice. That is the chapter in miniature: the review count is not what sets the price.

What crystal listings don't say

About 1 in 5 listings of commonly treated stones says so. For citrine it's 8 in 100.

Many everyday crystals are made, heated, coated or dyed. Most citrine is heated amethyst. Aura quartz is quartz with a metal coating. Opalite and goldstone are glass. Bright pink and blue agate is dyed. None of that is a secret in the trade. The question is whether the buyer can read it on the page.

Disclosure is a shop habit. Of 1,020 shops with at least three listings of these stones, 52% never say it, 44% sometimes do and 4% always do. And it isn't the premium shops leading: the highest-priced quarter of shops disclose on 12% of listings, the lowest-priced quarter on 27%.

Does saying it cost you?

Less than you'd fear. Citrine listings that say "heated" ask about 0.56 times as much as ones that don't, but heated and untreated citrine are different goods, so that's not a penalty for honesty. For aura quartz and moldavite, the wording makes no measurable difference. The premium sits with the word "natural": citrine with "natural" in the title and no mention of heat asks 1.31 times more.

Here is what good looks like. Heavenly Crystals Online and Throwin Stones state the coating on about 99% of their aura quartz. Pierres du Monde does it on 97% of its citrine. Rock Paradise labels 94% of its bright agate as dyed. Wholehearted Crystal Creations puts it in five words: opalite is a man-made glass. One line, and it costs nothing.

The other side: we found 94 titles that call opalite, goldstone or aura quartz "natural" without saying what they are. Add listings that say "natural opalite" in the description, and it's 577.

Words that move a price, and words that don't

A weight in the title goes with +87% even inside the same shop. "Healing", "gift" and "natural" go with nothing at all.

Because the shop is half the price story, we tested every listing feature twice: against other shops, and against the same shop's own listings. A word that only works across shops is a sign of the kind of shop that uses it. A word that still works inside one shop is closer to the real thing.

Size words are the strongest signal because they describe the piece: "XL", "giant", "large", "mini". "Rare", "one of a kind" and grade words like "AAA" still hold up inside a shop. "Genuine" and ALL CAPS look strong across shops and collapse to nothing inside one: they mark a type of shop, not a better listing. A healing claim goes with a 34% lower price across shops but only 10% inside one. It marks budget shops, not cheaper stones.

The honest part: many of the best words describe a bigger or rarer piece, not a better listing. When we paired a rich listing with a bare one of the same stone, shape and size inside one shop, the rich one asked 1.13 times as much, and that's within the noise. A long title won't add a premium to an ordinary stone. It does stop your stone being compared with a $2 chip.

From one cent to $98,500

The typical listing asks $55. The average asks $269, because the priciest 1% of listings carry 35% of all the dollars asked.

Most price reports describe the middle. The edges are where the stories are.

Same photo, same price

When shops sell the identical product, they agree. We found 131 groups where the exact same product photo appears in three or more shops, usually a supplier's stock image. In half of them the dearest shop asks no more than 1.23 times the cheapest. Listings that only share a bare name like "fluorite" spread ten times. Most of the price chaos is a naming problem: different pieces under one word.

Rare, common and odd

Painite, one of the rarest gems on earth, has 4 listings. Red beryl, jeremejevite and serendibite have one each. Moldavite, from one corner of central Europe, has 3,057 listings in 539 shops. Price per gram runs from half a cent (a 20 kg box of selenite logs at $100) to about $3,930 (a 61.77-carat opal). The heaviest single piece is a black tourmaline log of about 2.3 tonnes at $50,000.

About 1 listing in 12 sits ten times above or below the typical price for its own stone and shape. Some are museum pieces. Many are bulk lots, per-gram prices or placeholders: 100 listings ask under 10 cents. Before you copy a competitor's number, check its size, its unit and whether it's the same item.

If we opened a crystal shop tomorrow

Stonebridge's twenty-four years of selling crystals wholesale to shops, and one very large spreadsheet, point the same way. Price is not where a small shop wins. Who you are to the buyer is.

One more: don't copy a price without copying the piece. Identical items agree within about 1.2×. If a competitor is half your price, check it's the same size, grade and unit before you move.

How we did this

Data. We searched Shopify's public product catalogue the way a buyer would, in four regions and five price bands. These are asking prices, not sold prices. We then re-read a sample of product pages and shop homepages for review apps, review counts, social links, store size and full size-and-price options.

Processing. Stone, form, size, weight and wording were read from titles, variants and descriptions by rules. Prices are in US dollars at fixed rates (CAD 0.72, AUD 0.66, EUR 1.08, GBP 1.27). A sample of photos was checked with an image model and by hand.

How accurate is it? We checked 1,000 listings one by one

We drew 1,000 listings at random, one per shop, balanced across regions and price levels, opened each live product page and checked every field by hand. Here is how often our software got it wrong in the listings our analysis uses:

Errors are worse in EU and Canadian listings, where titles are often not in English. Prices held up: 83% of live prices matched our data within 2%. We checked the disclosure numbers separately against 250 full product pages: citrine came out at 8.2% (our data said 8.3%), opalite and goldstone at 34%, and aura quartz higher than we first measured, about half. We re-ran the main test on the 82,339 cleaner listings that name their stone and form in the title: the shop alone still explains 40% of price against 30% for stone, form and size together.

What that means for you: the shop-versus-stone finding, the disclosure gaps and the price spreads are solid. Size curves and the effect of words like "natural" and "healing" are a direction, not a measurement; noisy sizes tend to flatten the curves, so real spheres probably climb even faster than we show.

Limits. Each search returned similar numbers of listings per price band, so expensive items are over-represented. Averages lean high, which is why we lead with ratios and spreads; re-weighting moves the overall median from $55 to between $20 and $35 and leaves the ratios within a few percent. Everything is correlation. Stonebridge Imports and its retail shop The Rock Space are in the data like any other shop. No shop is named in a negative example, and shop names are hidden in every screenshot.

Questions shop owners ask