Wine-Searcher integrating LWIN codes. Will this finally fix the missing wines issue?

My biggest gripe with Wine-Searcher is that there seems to be no effective way to request corrections to their database. It’s easy enough to report pricing/offer mistakes, but reporting missing bottlings is almost impossible.

Yes. I’ve emailed them to no avail. For example, I once told them that the basic Mosquita Muerta Malbec was not listed. In response, one of their CRM Specialists emailed back saying that yes, of course it’s visible, pointing to this wine from the same producer, called Mosquita Muerta Blend de Tintas, which is, of course not a Malbec as its name clearly states. I couldn’t even attribute that one to her not understanding Spanish because “blend” is not in Spanish.

(CellarTracker’s opposite solution, simply allowing people to add wines, does cause listing proliferation, but Eric’s staff is quick to correct those when one flags it for them.)

So, it was with some hope that I read yesterday’s announcement that Wine-Searcher has incorporated LWIN codes to its database. At last, I thought, a system that can distinguish Fass/Fuder numbers for otherwise-identical German wine names (a persistent problem I had noticed for many wines) and can also quickly get products into Wine-Searcher. And it does the first, to a point.

You find an LWIN code here. And, if they have it --for example, you get LWIN 1672965 for Weingut Max Ferd. Richter, Brauneberger Juffer Riesling Kabinett Fuder 4-- you enter that string into WSP and out comes a perfect listing, different from the standard Juffer Kabinett (LWIN 1090918). Alas, my happiness was short-lived.

The first sign of trouble came when an LWIN code popped-up for a Weingut Max Ferd. Richter Brauneberger Juffer-Sonnenuhr Riesling Kabinett (LWIN 1266966), a wine that, as far as I can find, has never existed. There used to be a Juffer-Sonnenuhr Kabinett trocken some time ago, but that’s not the same thing.

No luck with Dr. Hermann Kabinetten either, for which LWIN has 2 entries, instead of at least 8. Wine-Searcher actually has 6 of those 8 already listed. Why didn’t they flag those 4 missing entries for Liv-ex (which owns LWIN) when doing the integration?

And the issue isn’t limited to Germany of course. LWIN doesn’t distinguish, for example, between Cafaggio Chianti Classico Riserva (no LWIN entry), and Villa Cafaggio Chianti Classico Riserva (LWIN 1106400), even though the former, a higher-end offering, has been on the market for 9 years and is not an obscure wine.

But in all of those cases the problem is a missing LWIN code. My saddest discovery was that the existence of an LWIN code for a wine did not translate automatically into a Wine-Searcher entry. I looked for a wine that has always been a PITA to find in WSP: Pierre Brisset Nuits-Saint-Georges Premier Cru Aux Thorey (LWIN 2239859). But I entered its LWIN code into WSP and got nothing (at least if I enter the words I get some offer listings). So it seems that, when doing the integration, not only did they not flag missing LWIN codes for wines with entries in Wine-Searcher, they also let existing LWIN entries dangle without pairing.

There must be a better way to do this.

LWIN covers about 150,000 wines. That is cute. CT has 4 MILLION wines in our database.

LWIN is/will be a nice Rosetta Stone for sites/retailers to at least synchronize their ability to “transact” on the core collectible wines. Not much more at this point.

Dear Guillermo

I’m sorry if there was a misunderstanding where Wine-Searcher could not match a ‘Malbec you told us was not listed’. If we cannot find a direct match to a specific wine listed in our database or one that is listed, but no online shop has for sale today, we do suggest in our results other wines to view from the same producer eg. ‘also from producer ‘Bodega Mosquita Muerta’. Requesting Wine-Searcher to seek out an update to a missing bottling is straightforward and uses the same email process as reporting a pricing error. The more detail one can provide the better. Ideally, a photo of a label or even a screenshot taken from a phone can prove useful as producers do not automatically share their labels with Wine-Searcher; so it helps to have hard evidence for a Wine Specialist to work with or go back to a producer with to confirm a bottling.

Regarding Liv-ex’s LWIN wine codes. As Eric mentions in his post below, LWIN numbers, while convenient to use, are a small subset of the full wine market. Such is the complexity of nomenclature in winemaking that any abbreviated code will be at disadvantage compared with the in-full wine name when it comes to matching a wine search by an individual to a product’s given name. In Wine-Searcher’s case, we also simultaneously match the above two names with the myriad of different names that stores use to describe the same product. Plus we pick up additional purchasing and investment cues such as OWC, damaged label, etc. that differentiate one shop’s offer from another. Last night’s update of names, descriptions, and prices was in the millions. It is a Herculean task - one that we are passionate about.

The term ‘integration’ may have laid the grounds for some confusion. Wine-Searcher software now reads LWINs. LWINs are read-only. They are not read/write. The issuing and correction of an LWIN is undertaken entirely by Livex.

I did that. In my email, I sent a link to the wine.com offering (which includes a picture of the label). That’s why the response was not satisfactory. But thanks for taking the time to reply.

I appreciate the clarification. And I do grasp the complexity and the awesome nature of the task. That’s part of why I thought the LWINs would help. I just think there must be an easier way for all parties than an email back and forth for each missing wine. I’m not sure what the solution is. But I think the way CellarTracker does it is better, even if it’s not perfect either (but then what is). And if it must be an email, then at least a presumption that if someone is emailing with specifics about a missing wine, unless they’re unambiguously wrong, they’re probably right, because why else would anyone email.

Thanks again for replying.

Seems like an opportunity for CT. Unique identifier for every wine, with CT+community error correction. Eric, what do you think? Of course, no clue how you monetize this but there is always a way for good data to create value for someone.

Related - it would sure be nice if wine producers created canonical sets of entries in CT for their wines. If they did, and CT assigned unique IDs during this process, it would solve the problem of creating the dataset, and also improve CT data usage/quality. Again, monetization is another question.