Vendors and consumers: what would you think about having a (PS?) centralized colored stone database?

[deorwine]

Just for kicks -- we were tossing around ideas in this post about a centralized colored stone database like there is for diamonds. @Garry H (Cut Nut) started it off, and I thought it was kind of interested so I kept the discussion going. Lots of interesting points made in that post. But anyway, just because I like doing thought experiments and seeing what people think about things...

So, the question: what do you think about having a centralized colored stone database on PS like the one that exists for diamonds? Of course it would have to be different, have a lot more color options, etc. That is also great fun to discuss (and I got a new app out of it, thanks @PrecisionGem ) and feel free to do it in this thread if you want.

I've set the poll so that people who are both sellers and consumers can vote twice. I wasn't really thinking about second-hand private sellers like on LoupeTroop (I assume these are not present in the PS diamond database) but use your judgment as to whether you think that counts

 

[chrono]

While I would love standardization of colour and a database, I worry that only the huge companies or middle person will benefit and small cutters will be left out. My best experience has been a mix of both (example Prima for the large guy and Kingdom of Red Gems for the unknown guy).

It will only benefit larger or more expensive gems and smaller or gems less than $500 will not be worth listing, from the perspective of the seller.

 

[deorwine]

*nods* A couple of people brought that up in the other thread as well. I wonder if something like this would lead to a bifurcation of the market, with expensive gems dominated by huge companies and cheaper gems by small ones. That would be a shame, as I like cutters being able to cut both cheap and expensive gems, and not being restricted.

I'd love to hear more from vendors -- do all of you guys agree that small cutters would be left out? (I know @PrecisionGem does.)

While I'm tossing out crazy ideas, what about the converse? What if vendors signed up for a database of consumer requests -- like, I could put in a request for "slightly purplish red spinel, slightly included, strong saturation, 0.5-1ct, precision cut, up to $1000" and have vendors in the system email me with what they have in stock. I suppose the problem with that is that you'd probably get a lot of spamming by people who want vivid 4ct gems for $100... there'd have to be some way to filter out obviously unfillable requests, I think.

 

[minousbijoux]

If it would work, it would be great. But can't imagine it would. Like herding cats.

 

[prs]

You will need to have verified and color graded stones in your database for it to have as much consumer value as the diamond database. I believe AGL is the only US lab that offers color grading in their "Prestige Grading Report". I don't know if vendors will be willing to pay the high cost for that report, the problem being that most consumers will have no idea how to interpret AGL's color grading system. Here's a copy of the AGL price list.



FWIW the top auction houses almost always provide AGL certs for their colored stones. I have yet to see one that included the color grading.

 

[arkieb1]

Here is another issue I've had. Say @Garry H (Cut Nut) creates a data base and a number of vendors sign up to use it. To make it easy it has a number of different colour shades vendors can pick and auto descriptions they can use to describe the colour of each stone.

No one in the industry can agree exactly on what things like "cornflower blue" and pigeon's blood red looks like. There is a wide variation for example between blues that look like real cornflowers ie a lighter to mid tone all the way to a fairly dark electric blue tone described by vendors as "cornflower" blue. Similar there is a wide variation from a deeper darker red to a vivid bright red as pigeon's blood red.

The industry itself cannot agree on colours. So all I see is one vendor picking one colour swatch as representing the exact colour or tone of the stone they are selling and other picking an entirely different swatch for a similar coloured stone....

Even picking swatches and ticking boxes when it comes to colour is subjective. I really like the idea @Garry H (Cut Nut) had about master colour sets like diamonds have but the skeptic in me just cannot see vendors with the time or the desire to accurately use them.

So if the stones in the data base all have inaccurate or different descriptions of similar colours and the listings are still based upon the subjectivity of the vendor then the whole thing becomes moot.... the only way it works is if everything has centralised certificates like diamonds do by recognised bodies like GIA and AGL and at a cost of up to several hundred dollars per stone that excludes the viability of cheaper stones and means everything will go up in pricing.

 

[Garry H (Cut Nut)]

Here is another issue I've had. Say @Garry H (Cut Nut) creates a data base and a number of vendors sign up to use it. To make it easy it has a number of different colour shades vendors can pick and auto descriptions they can use to describe the colour of each stone.

No one in the industry can agree exactly on what things like "cornflower blue" and pigeon's blood red looks like. There is a wide variation for example between blues that look like real cornflowers ie a lighter to mid tone all the way to a fairly dark electric blue tone described by vendors as "cornflower" blue. Similar there is a wide variation from a deeper darker red to a vivid bright red as pigeon's blood red.

The industry itself cannot agree on colours. So all I see is one vendor picking one colour swatch as representing the exact colour or tone of the stone they are selling and other picking an entirely different swatch for a similar coloured stone....

Even picking swatches and ticking boxes when it comes to colour is subjective. I really like the idea @Garry H (Cut Nut) had about master colour sets like diamonds have but the skeptic in me just cannot see vendors with the time or the desire to accurately use them.

So if the stones in the data base all have inaccurate or different descriptions of similar colours and the listings are still based upon the subjectivity of the vendor then the whole thing becomes moot.... the only way it works is if everything has centralised certificates like diamonds do by recognised bodies like GIA and AGL and at a cost of up to several hundred dollars per stone that excludes the viability of cheaper stones and means everything will go up in pricing.

I have sent a supplier in Thailand a small glass lid box with several different cheap paste/synthetics of all colors. When we are communicating color we/he sits the gem on the box and takes a photo.

My long time friend Dave Atlas (aka OldMiner) and I have been discussing making sets of shallow, ideal and áverage' or slightly deep CZ's in a little box. I could sell those on Ideal-scope.com for around $15-20 in a nice hinged glass lid presentation box.

Color master sets would be way cheaper.

 

[Garry H (Cut Nut)]

Regarding cost. PS is a small business. A micro business. We have less than 20 companies pay us money and less than a dozen who list any diamonds because most diamonds are picked up from Indian manufacturers data bases by perhaps 100,000 web sellers around the world.

We do not want more than a few top service vendors selling the same stones.

There are no such shared colored gem data bases.
There would be room for maybe hundreds of vendors filling all sorts of market niches - they need not pay much to list.
But what we have that other eBays and Etsy etc do not have is this vibrant smart savvy community to keep vendors honest.
I imagine we might end up with a few dozen types of vendors specialising in gem varieties and specialist high end rare certified goodies. Each catagory might have 3 to a dozen vendors who find it works for them.

 

[qubitasaurus]

I am a bit hesitant to post this, but is that the colour master set? As I don't think I would buy anything that matched any of those colours, except possibly a rubelite tourmaline with the colour of the stone in the bottom left hand corner. Otherwise those are definitely not what one typically wants to buy in coloured stones. None of them are glowy and none of them are vivid. I could see myself using that as a benchmark (but items would need to signifcantly exceed that benchmark for me to consider them). Maybe it could be possible to build several master sets of this type, each at an increasing quality level. And then post a pic against the set that best matches the stone (while noting which set that was). Under some standardized lighting conditions. I acknowledge that if something like this was done then it could indeed make it easier to compare different photos from different vendors.

You know the lot of us are still going to ask for a picture on the back of someone's hand in daytime indoor lighting, and dim flourescent nighttime lighting though . Coloured stones shift so much that one shot in one lighting usually tells you little about how the gem is actually going to look for most of the day.

 

[Garry H (Cut Nut)]

I am a bit hesitant to post this, but is that the colour master set? As I don't think I would buy anything that matched any of those colours, except possibly a rubelite tourmaline with the colour of the stone in the bottom left hand corner. Otherwise those are definitely not what one typically wants to buy in coloured stones. None of them are glowy and none of them are vivid. I could see myself using that as a benchmark (but items would need to signifcantly exceed that benchmark for me to consider them). Maybe it could be possible to build several master sets of this type, each at an increasing quality level. And then post a pic against the set that best matches the stone (while noting which set that was). Under some standardized lighting conditions. I acknowledge that if something like this was done then it could indeed make it easier to compare different photos from different vendors.

You know the lot of us are still going to ask for a picture on the back of someone's hand in daytime indoor lighting, and dim flourescent nighttime lighting though . Coloured stones shift so much that one shot in one lighting usually tells you little about how the gem is actually going to look for most of the day.

It is just an example - I put together tow sets for one supplier with a few cheap stones.
The point is when we both have the same set a photo taken on it shows how the stone looks.
We could easily buy 1,000 of 10 different color synthetics for peanuts and put together sets.
It is just an idea for discussion. Not a plan

 

[qubitasaurus]

It is just an example - I put together tow sets for one supplier with a few cheap stones.
The point is when we both have the same set a photo taken on it shows how the stone looks.
We could easily buy 1,000 of 10 different color synthetics for peanuts and put together sets.
It is just an idea for discussion. Not a plan

Sounds good. I agree you could make many such sets. And run the gamut of qualities -- as synthetics can be very pretty.

But actually as I have been thinking about it I have realized, you just validated what Gene was saying to you earlier .

Your boxes are all about creating a joint reference which both halves of the transaction had in front of them. So when you talked both could look down at this exact common reference point sitting in front of them.

To replicate the same thing with a consumer and a vendor you should use gemewizard. It is free so both buyer(/here consumer) and seller(/vendor) will both be looking at the exact same tableaux sitting right in front of them as a reference point for the discussion. This would be the closest you could get here to 'your shared duplicated gem box'. Because it is not that realistic to have consumers have one of these reference synthetic gem boxes in front of them while talking to the vendor. But they could all put gemewizard on their phones. And we could take pictures next to such a tableaux. To show the comparison.

You will still have to deal with the request for at least one picture in diffuse daylight, one in indoor fluorescent nightime lighting and one in direct sunlight on the back of someones hand, though. This is because some gems shift colour so much that they can look blue green or salmon pink in one lighting. And look like muddy water in another. To be honest I have had gemewizard on my phone for a while. But I dont bother much with it, as I am more interested in this info (as I really dont like stones that look like muddy water. So this will be the first thing I will be checking for. Actually this is one if the most important key search criteria for me. Which would also be a very difficult one to quantitatively evaluate in some standardized system. The other important ones would would be vividness of the colour. And cut issues such as windows, bowties, extinction, etc. Again I know it immediately when I see it. But I dont know how youd quantitatively measure it in an unbiased/standardised way in order to construct a database).

But yes I can see some possibility for the standardization to do some good. I just think Gene may be ahead of the curve . I guess maybe I could be cautiously optimistic.

 

[cm366]

Great idea to offer the service. Might not work out (for a reason enumerated above, on the previous thread, or any number of others). But won't break the site (or any existing vendors' models) unless it completely remodels the CS trade for the better.

And...it might help do just that!

 

[jordyonbass]

No one in the industry can agree exactly on what things like "cornflower blue" and pigeon's blood red looks like. There is a wide variation for example between blues that look like real cornflowers ie a lighter to mid tone all the way to a fairly dark electric blue tone described by vendors as "cornflower" blue. Similar there is a wide variation from a deeper darker red to a vivid bright red as pigeon's blood red.

The industry itself cannot agree on colours. So all I see is one vendor picking one colour swatch as representing the exact colour or tone of the stone they are selling and other picking an entirely different swatch for a similar coloured stone....

This is especially true for the Opal world where we see so many different colours, patterns and body tones. As an example of this, the body tone category seems to be one so many vendors have trouble with. While there are body tone scales that are supposed to set an objective grade for Opal, I have seen long-time vendors selling stones as Black Opal that I would sell as a Dark Opal or Semi-Black Opal. I've also seen some vendors selling Dark Opals that look an obvious N2-3 body tone to my eye.

I think the issue for Opal is that if you were to put out a request and stated that you want an Opal with a bunch of different requirements - you're going to get stones that all look completely different from the vendors that have something in stock that matches your criteria.

 

[Garry H (Cut Nut)]

This is especially true for the Opal world where we see so many different colours, patterns and body tones. As an example of this, the body tone category seems to be one so many vendors have trouble with. While there are body tone scales that are supposed to set an objective grade for Opal, I have seen long-time vendors selling stones as Black Opal that I would sell as a Dark Opal or Semi-Black Opal. I've also seen some vendors selling Dark Opals that look an obvious N2-3 body tone to my eye.

I think the issue for Opal is that if you were to put out a request and stated that you want an Opal with a bunch of different requirements - you're going to get stones that all look completely different from the vendors that have something in stock that matches your criteria.

My friends Andrew and Damien Cody (other famous Aussies, we don't just have movie stars and rock bands hahaha) developed an opal grading system. Andrew was also El Presidento of the International Colored Gemstone org for a few years. I am sure they could share some opal stuff.

 

[Garry H (Cut Nut)]

e.g. https://www.syntheticgems.org/
I see lots of these companies at the HK show

 

[Arcadian]

Sets of basics? Ok sure. but for all stones? no. that gets ugly real fast IMO.

Garnets, tourmaline, spinel, zoisite..... those get weird and in the case of tourmaline the color combinations, the open and closed C axis, different colors on ANY axis, and this is whats hard to have a "standard". Hell fluorite gets weird. And then you have the gemstones like lapis, turquoise, rhodochrosite...and even titanite.

Then theres discussion on undetectable treatments. For instance, you can have a blue topaz thats untreated and a blue topaz that is treated. Are you going to be able to tell? No. And irradiating the stones is typical and normal.

The other side of it is to not make colored stones try to fit into the wrong pegged hole. People have some weird expectation that what holds true for diamonds holds true for all gemstones and that is really not true at all. For instance, emeralds, have jardin typically. thats normal for that type of stone and you'll see it unless the stone's been treated. there's not that much jardin free emerald coming out of anywhere (if it does its gonna be megabucks!) Then there's the treatments that are acceptable and what TYPES of treatments are actually acceptable and the LEVEL of treatment.


Just my thoughts but real education has to take place along with tempering real expectations if you're going to consider building this.

 

[Garry H (Cut Nut)]

You are right @Arcadian and others. I am over thinking it.

Now FYI - if we do go ahead with this crazy idea - It took Andrey about 18 months (from memory) to build the jewelery search machine, and he has lots of projects lined up. Remember PS is him, Ella, a couple of part time blog writers and contract developers. PS has never once paid a dividend.
But I think this could work and be beneficial for gem collectors, high end sellers right down to helping people learn how to by better gems in B&M's and HSN and Costco etc.

But my eventual goal would be to vastly improve the average beauty of the average gem, which is what I have been doing on PS for diamonds for 21 years.

 

[Arcadian]

@Garry H (Cut Nut) I think its best to start small. Colored gems is a giant area to wrap ones head around....just non stop education for many of us deep into it.

So if I had to build something like this, I'd start with the money stones.

The ones that would be worth time would be people who are looking for engagement rings/anniversary rings, push presents, special birthday gifts

Sapphires, ruby, red and pink spinel, tsavorite garnet, paraiba tourmaline.

Most gem cutters sell sapphires and lets be honest, no one wants an ugly or badly cut one as an engagement ring, and lots of people don't want "blood diamonds" but want an equally bloody sapphire.

So just my thought; start single category and go with the basics in sapphires; blue, yellow, pink. pads maybe but they're ridiculously expensive, rarely precision cut, color can be an issue.

Then add rubies if that works out well because then it could be easy enough to do. In my view, really nice rubies are rare, they're really not seen a lot in the US as in other places. The hardest part of rubies is that depending on where they're sold, the report can say ruby and still be what many consider to be a vivid pink sapphire. Acceptable Treatments can be an issue but thats with most of these stones.

If you chose to focus on Spinel, focus on Pink and Red only as a starter because you have several locales just under these 2 colors.

Garnets. Tsavorite garnets only because they were made well famous by Tiffany's. Spessertine garnets have fallen out of high favor, and mahenge garnets are in. They may be out next year you never know.... I honestly wouldn't get too deep in the paint on the other garnet types right away. Tsavorite from what I've read thus far seem to be pretty impervious to treatments.

If you wanted to do tourmaline, don't do them all. but also know that the money stone is usually not precison cut. Thats the Paraiba.

Even on the stones I mentioned there's lots to know about treatments but this should be fairly easy info to gather as there's scads on this forum already...


I wouldn't bother with cheap stones. Almandine garnet, Topaz, amethyst, peridot, citrine...thats what you see at costco and the like and can be seen as cheap unless some brand name is attached to it... and quite honestly very little of it will be precision and they really don't care if it is, cheap as chips.....

 

[chroman]

Given the discussion about how to get “good data” into such a system, what if we turn the problem around..

Lots of folks here have looked at lots of vendor photos online - could that be leveraged?

The system then doesn’t need any new inputs from vendors - it’s becomes a listing aggregator and rating voting system. Something along the lines of the “someone’s got to buy this” thread, but with more automation behind it.

When new listings appear, PSers get prompted to rate the listing on a few dimensions.

 

[Garry H (Cut Nut)]

Given the discussion about how to get “good data” into such a system, what if we turn the problem around..

Lots of folks here have looked at lots of vendor photos online - could that be leveraged?

The system then doesn’t need any new inputs from vendors - it’s becomes a listing aggregator and rating voting system. Something along the lines of the “someone’s got to buy this” thread, but with more automation behind it.

When new listings appear, PSers get prompted to rate the listing on a few dimensions.

Some of our Diamond experts say they can tell the cut quality and inclusion visibility on different images and videos from different set ups.
Is that what you kinda mean chroman?

 

[Garry H (Cut Nut)]

@Garry H (Cut Nut) I think its best to start small. Colored gems is a giant area to wrap ones head around....just non stop education for many of us deep into it.

So if I had to build something like this, I'd start with the money stones.

The ones that would be worth time would be people who are looking for engagement rings/anniversary rings, push presents, special birthday gifts

Sapphires, ruby, red and pink spinel, tsavorite garnet, paraiba tourmaline.

Most gem cutters sell sapphires and lets be honest, no one wants an ugly or badly cut one as an engagement ring, and lots of people don't want "blood diamonds" but want an equally bloody sapphire.

So just my thought; start single category and go with the basics in sapphires; blue, yellow, pink. pads maybe but they're ridiculously expensive, rarely precision cut, color can be an issue.

Then add rubies if that works out well because then it could be easy enough to do. In my view, really nice rubies are rare, they're really not seen a lot in the US as in other places. The hardest part of rubies is that depending on where they're sold, the report can say ruby and still be what many consider to be a vivid pink sapphire. Acceptable Treatments can be an issue but thats with most of these stones.

If you chose to focus on Spinel, focus on Pink and Red only as a starter because you have several locales just under these 2 colors.

Garnets. Tsavorite garnets only because they were made well famous by Tiffany's. Spessertine garnets have fallen out of high favor, and mahenge garnets are in. They may be out next year you never know.... I honestly wouldn't get too deep in the paint on the other garnet types right away. Tsavorite from what I've read thus far seem to be pretty impervious to treatments.

If you wanted to do tourmaline, don't do them all. but also know that the money stone is usually not precison cut. Thats the Paraiba.

Even on the stones I mentioned there's lots to know about treatments but this should be fairly easy info to gather as there's scads on this forum already...


I wouldn't bother with cheap stones. Almandine garnet, Topaz, amethyst, peridot, citrine...thats what you see at costco and the like and can be seen as cheap unless some brand name is attached to it... and quite honestly very little of it will be precision and they really don't care if it is, cheap as chips.....

Arcadian - it is not complex to allow vendors to list whatever they want. It's just a detailed menu. And if we need sub menus etc then Andrey and the developers can work that out.
There will be othr sellers with menus worked out and we can copy and include / exclude etc.

 

[chroman]

Some of our Diamond experts say they can tell the cut quality and inclusion visibility on different images and videos from different set ups.
Is that what you kinda mean chroman?

Maybe, but I think what I’m picturing is even simpler.

The average query in CS seems to be, “help me find the best $SHAPE $COLOR $SPECIES thats in my $BUDGET!”

At which point, PSers manually comb inventory listings and post their favs.

So what if we help accelerate that query. Reliably scraping shape and color from listings will be tough, and may require an initial manual labelling step by someone.

Then, give PSers a “rate these stones” option, where they are shown listings they haven not seen already and can given them 1 of 3 labells
- I would buy this stone
- I wouldnt buy this stone, but i’d recomment someone else check it out
- I would not buy or recommend this stone

Those ratings get rolled up into the measure of “best” for your query.

Getting price for the query can be tricky if its not posted. You may be able to estimate a range from similar stones on the market, but that needs a distance metric (more human ratings could work here too..)

 

[PieAreSquared]

Lots of folks here have looked at lots of vendor photos online - could that be leveraged?

When new listings appear, PSers get prompted to rate the listing on a few dimensions.

This sounds good except for the very real possibility that the vendors with the best photos, not necessarily the best stones, will get voted up.
How could that be overcome?

 

[chroman]

This sounds good except for the very real possibility that the vendors with the best photos, not necessarily the best stones, will get voted up.
How could that be overcome?

Yeah, I was assuming this would be handled by the collective implicit mental calibration of vendors photos to reality. But that might be a big assumption.

 

[Garry H (Cut Nut)]

Maybe, but I think what I’m picturing is even simpler.

The average query in CS seems to be, “help me find the best $SHAPE $COLOR $SPECIES thats in my $BUDGET!”

At which point, PSers manually comb inventory listings and post their favs.

So what if we help accelerate that query. Reliably scraping shape and color from listings will be tough, and may require an initial manual labelling step by someone.

Then, give PSers a “rate these stones” option, where they are shown listings they haven not seen already and can given them 1 of 3 labells
- I would buy this stone
- I wouldnt buy this stone, but i’d recomment someone else check it out
- I would not buy or recommend this stone

Those ratings get rolled up into the measure of “best” for your query.

Getting price for the query can be tricky if its not posted. You may be able to estimate a range from similar stones on the market, but that needs a distance metric (more human ratings could work here too..)

We would need to be careful of shilling in such a setup chroman.

 

[Garry H (Cut Nut)]

This sounds good except for the very real possibility that the vendors with the best photos, not necessarily the best stones, will get voted up.
How could that be overcome?

A simple concept called competition.
like keeping up with the Jones's

 

[Lilith112]

Well, another issue is that a lot of vendors-- especially wholesalers-- are shifting towards social media platforms. For example, Chinese vendors post majority of their gems through WeChat-- but you need to be friends w/that person on WeChat to see their posts. Japanese vendors primarily on Line Chat while a lot of businesses only advertise through WhatsApp.

I like the idea of a database in theory, but in reality, how are you going to scrap for WeChat vendors? Taobao-based vendors? 闲鱼vendors? I just don't see it working well outside of North America/Europe sellers.

 

[Garry H (Cut Nut)]

Well, another issue is that a lot of vendors-- especially wholesalers-- are shifting towards social media platforms. For example, Chinese vendors post majority of their gems through WeChat-- but you need to be friends w/that person on WeChat to see their posts. Japanese vendors primarily on Line Chat while a lot of businesses only advertise through WhatsApp.

I like the idea of a database in theory, but in reality, how are you going to scrap for WeChat vendors? Taobao-based vendors? 闲鱼vendors? I just don't see it working well outside of North America/Europe sellers.

I NEED TO BE CLEAR - WE WILL NOT AND NEVER WILL SCRAPE INFO FROM OTHER SITES. Sorry for shouting, caps were on - but we do not do that!
Vendors use an automated .csv file upload which Andrey has helped them automate.

 

[Lilith112]

I NEED TO BE CLEAR - WE WILL NOT AND NEVER WILL SCRAPE INFO FROM OTHER SITES. Sorry for shouting, caps were on - but we do not do that!
Vendors use an automated .csv file upload which Andrey has helped them automate.

That...still doesn't really get at what I pointed out lol, namely that the database would end up being very North America/Europe centric. While this can be fine given that majority of PS seems to be based in those regions, most CS wholesalers are Asia-based anyways. And most of them are only reachable through closed social media platforms.

Just my two cents based on my personal experience. While there's a few Asia-based vendors who would probs be interested in joining the database, international shipping is a major hassle for others, especially given how much no-hassle money they can make in the China/HK markets.

 

[Garry H (Cut Nut)]

That...still doesn't really get at what I pointed out lol, namely that the database would end up being very North America/Europe centric. While this can be fine given that majority of PS seems to be based in those regions, most CS wholesalers are Asia-based anyways. And most of them are only reachable through closed social media platforms.

Just my two cents based on my personal experience. While there's a few Asia-based vendors who would probs be interested in joining the database, international shipping is a major hassle for others, especially given how much no-hassle money they can make in the China/HK markets.

Firstly almost all the virtual diamonds listed on Pricescope are actually still in India and ownwed by big manufacturers. When a consumer anywhere in the world orders them the owner ships to the seller or often drop ship directly to the consumer.
Secondly, when I go to trade fairs there are some gem and pearl wholesalers who sell to me prior to the show starting from Hotels etc and do not want their wholesale clients who are exhibiting at the show and paying for stands etc to know that you are buying directly from their suppliers.
This is normal as one guy might specialise in Sri Lankan goods and loans his goods to 2 or 3 sellers who have an entire range of goods from 20 sources.
some of each type of seller might be interested in listing their goods on PS.
Remeber too that you can find the same diamond from a few sellers on PS.