Creating the Ultimate Rum Database – Part 2: RumX

 1 Response »
Sep 212022
 

Part I was an extended discussion on the evolution and development and early efforts to create a comprehensive data set – a database – of rums. Not just current rums available now, but all of them, from all times, all eras, all countries. That it has never been done doesn’t make it any the less important…or dreamed of.1

To be sure there were some resources available, as listed – the problem was they were scattered, inconsistent, incomplete or did not have such a focus, and almost all that did gradually went dark because of the effort of maintaining them. Peter’s Rum Labels and Rum Ratings remain the best ones that are still live, yet they have limitations of their own. Sites like the American Rum Index and Australian Rum Index remain too small-scale in scope and rarely list rums at all; and review sites with real quantities of reviews — WhiskyFun (the champ with 1600+), the Fat Rum Pirate and this site (nearly a thousand apiece) — can only list rums that have been tried and to which there is access; still other sites like Barrel Aged Mind, Barrel Aged Thoughts and Single Cask Rum (which list issued rums as part of company profiles) only focus on what interests them, and are hardly exhaustive — and therefore there are gaps and fissures all over the place.

The time was therefore ripe for the next phase in such rum databases’ evolution: crowdsourcing and the mobile app. Already technology was moving to a point where solutions in other areas (where large volumes of data had to be collated) were seen as potentially applicable here, most notably that of setting up easy-to-use online infrastructure and letting the users provide the data.

Given the user-driven examples of Wikipedia and Rum Ratings in particular, the only real surprise is why it took so long for some enterprising rum aficionado to combine crowdsourcing with the increasingly headlong move towards portable devices and port the entire concept to mobile in the first place. Maybe it’s because app development is a young person’s game while the real deep diving rum chums are all old farts using desktops (like me, ha ha). The advantages of an app based on mobile rather than desktop (or laptop) technology are clear: in real time and usually on the fly, users provide the input, the content, the tasting notes, the scores, the production details, the label data — and the app or its owner acts as a middleman and moderator, curating the content for error checking, duplication and incorrect data. In this way the combined power of many creates a greater whole than any one person could possibly do alone and if crowdsourced edit and error correction gets folded in, well, you really have something here.

While Rum Ratings was the first application to take crowdsourcing seriously, it was geared for the larger screens of desk- and laptop, and hampered by the fact that the creator, for all the good intentions, was not so deep into the rum culture as to note the wave of consumer requirements and enthusiasm which would have made the website more useful. It was a hobby project that had great utility for a general rum drinking audience, but never stepped up to the next level to make it some kind of de facto leader in the field, or a standard of any kind.

The origin of the application that was and remains the closest to realising the vision of a publicly available, curated, and comprehensive rum database was actually not — then or now — created for that purpose. Like most programs, websites or applications that enthusiastic people slapped together in the past two decades with enthusiasm, gumption, gallons of coffee and too little sleep, its genesis was personal. Oliver Gerhardt in Germany was getting interested in rum, had a control spreadsheet to record his tasting impressions of those rums he had sampled, and a Master’s degree project to develop a mobile app: so he simply combined the two and came up with a small and very rudimentary application he called “Rum Tasting Notes”, which replaced his spreadsheet tasting diary.

Unsurprisingly for a first-gen effort where the basics and philosophy hadn’t been firmly nailed down yet, it was primitive. The database mechanism was clunky and at first very manually driven. The architecture lacked community ratings or a tasting feed, and there was only a very rough way to create tastings: if one wanted to record tasting notes it required manual fill of key data with no error checking. There was no true database in the backbone. The point-scale was between 1 and 10 with 0.5-point increments and thus covered only one fifth of today’s value range. But there were a few sliders, a scoring mechanism and a comment field, so the core features were already there.

After playing with it and expanding the facilities of the app for many months, he realised his pet project actually filled a niche that was imperfectly addressed by the existing sources available at the time. Perhaps others would be interested? In early 2018 he showed it to some friends and the app even as it was, was so well received that in December of that year he finalised the first version and uploaded it to the online app ecosystem, for free (it remains a free app as of this writing). Initially RTN was provided to the iOS app store, but an Android version was soon provided as well. At the time it had tasting notes — for that was the purpose of the app – for a total of 300 rums which had been cobbled together from his own notes, blogs, online stores and even books.

Although initially released only in Europe, several things caused the app to grow by leaps and bounds. For one, Oliver brought in several volunteers from the rum community to share it around, talk to fans about it — some, like Benoit Bail-Danel were well known and had good reputations in the field, so their word carried weight, and their social media posts and commentary accelerated the visibility and acceptance. Secondly, he added to the development team quickly: critically this was Jakob Schellhorn for product management and marketing, Marcus Rottschäfer for the recommendation engine and machine learning aspect, Vincent Kesel who played a major role in the development of the android version, Theresa Plos for her lovely design work, and other users and early adopters who provided useful feedback and acted as proselytisers for RTN. Clearly, then, the moment it went up as an app, commercial possibilities were being explored, and it was not going to remain a guy-in-a-garage thing forever.2.

And it didn’t, because it was constantly tweaked and redesigned as more functionalities were added, thought of or asked for. The name got switched from Rum tasting Notes to RumX in a major update in 20213. A website was set up for future expansions, where searches can be done. Both the recommendation engine and the tasting engine became more sophisticated; the latter allowed for both pre-defined words and manual fields; scores were made easier, and extra fields were added for new rums’ background details (a godsend to a guy like me). Error checking, data integrity and duplicate elimination was beefed up as rums were added, and continues to be scrutinized regularly (no rum gets added without being vetted first). And in mid-2022 the app was expanded to be available in North America.

Users could now, on the road and with minimum fuss and bother, add new rums and define where it was bought, basic details about it, whether it was a sample or a bottle, and how much it cost. Bar codes could be scanned in. Users could build a collection of their bottles, their tasting notes, their wish lists, interact with each other through a community feature, link to reviews and websites…the app has become something of a one-stop shop for all the things that individual sites once did – likes, discussions, bottle splits, shares, tasting notes, charts, scoring, and so on – while seamlessly integrating the experience and making it easier. Plus, it was mobile, so people could use it in real time when they were shopping, scan bar codes into it to get and upload data, check out aggregate scores on the fly and, as time went on, even check out prices and shop online for their rums. People in a shop used to have to search for bloggers’ reviews to see whether a rum might be worth buying – this app makes that an option, not a requirement, and it’s faster.

Of course, it’s not perfect: no app really can be and what one person likes might be an anathema to another. Adding a new rum on a smartphone takes time and is still something of a pain in the ass. Tasting notes icons for quick selection are not entirely intuitive (though the colour coding does help) and it’s not always clear what a slider or icon description might actually be for (e.g. “roasted”). It also lacks speed for multiple tastings and updates which a desktop version would assist. There are many like me who have hundreds of tasting notes and would like to add that to the database but it’s not feasible to do so. On a phone it’s just not going to happen: yet there is no facility to do so via an API upload or a desktop version of the program (as yet).

But circling back to the main theme of this two part essay, RumX’s influence and importance exceeds its user-friendliness and multi-functional abilities, to an extent not clearly appreciated by casual users, or even, perhaps, its creator. Most users, after all, just want to know whether to buy a rum or not, or what a bottle they have should taste like, or what others thought of it.

Whether by design or not, though, RumX has become far more than just a tasting notes diary, score aggregator, digital collection builder and rum collector’s app. It has become a central hub in the rum consumer’s ecosystem, connecting shops, reviews, scores, specifications, users and even conversations. And as time went on and more and more people adopted it and began adding to it, what it brought to the table was a digital, mobile tasting note app, reasonably easy to use, minimising the amount of typing and quick to update by the average user. Nothing we’ve ever seen has even come close to this kind of broad functionality. And this in turn has led RumX, not quite four years after its introduction, to already boast more than 13,000 rums in its database.

Just think about what that means. In 2013 I opined to a friend of mine that perhaps there might be 5,000 rums in the world. Ten years on, I know that it was a woeful understatement – because having seen Luca Gargano’s 5,000-rum warehouse and knowing of Steve Remsburg’s 1,200+ rum collection, and considering the hundreds or even thousands of new rums that get released by old and new bottlers, distillers and indies every year, I’m aware that we are in very real danger of just getting lost in the wash. It’s too much, too fast – no writer can keep up, not without help, time, sponsorship or funding – so to have this one resource that lists more than anyone else and has it available to everyone, is a huge benefit to the entire community.

For once, we have the facility – limited, but still more than before – to tap into the biggest single repository of rum bottling information that exists for the general public. At last we can tell if the Cadenhead VSG 73.6% 1990-2003 Guyana rum is real, or a mislabelled sample bottle, and not spend two days tracking it down. Individual reviewers and writers might deep dive into a single rum and put more info out there (and in better prose), and those who write informational pieces about distilleries, distillation, companies, personalities, styles, countries and so on will never be out of work. But in aggregate and for what it is, RumX is quite simply the biggest database and the best resource of rum bottling information out there that’s available and accessible to the general public. If Oliver and his team can find a way to upload bulk information or allow the inclusion and processing of data more rapidly by the reviewing sources who have been the backbone of the writing community for so long, I don’t doubt that RumX can top twenty thousand entries in another few years, easy.

And that serves all of us who want to know more about the more obscure older bottlings we occasionally run into, as well as the newest and bestest by the big names. Finally, it looks like after decades of trying, the ultimate rum database has been found: not in the basement labour of an unknown and unappreciated solitary rum lover who never shares because it’s never finished, but — as it almost had to have been — in our own collective consciousness, in each and every one of us who love the spirit.

Other Notes

  • I could have made this a single essay and just added RumX as the last entry in the list of databases, but that would have meant shortening the info I had on the app (hat tip to Oliver, who provided much of it) and I felt it to be useful in and of itself.

Sources

Creating the Ultimate Rum Database – Part 1: Early Efforts

 8 Responses »
Sep 182022
 

Since the very beginning of the distributed and engaged rumiverse, there have been movements — almost all by individuals — to catalogue all rums in existence. All of them came up short, failed, or were abandoned…though many, in hindsight, pointed to the desirable characteristics of some as-yet undeveloped system and encouraged the next generation of creators. Yet perhaps now we are on the edge of cracking the problem. This two part essay charts the beginnings of such projects, why they are important, and where it all seems to be leading.

For rum deep divers, researchers, auction houses, the curious, the writers, the inheritors of dusty bottles, for all these people and more, a good rum database – a listing of rums – is now needed more than ever before. A good database or website that catalogues rums would not only have technical details – producer, bottler, source material, distillation notes, dates, strength, age, additives, country or city or company of origin and so on – but link to secondary and tertiary sources, provide label photographs, list online review sites, available shopping sites, and have commentary. In today’s world where questions asking about this or that rum pop up all the time and with ever-increasing frequency, the importance of such a database cannot be casually dismissed. It can be used to gauge value, chart trends and identify purchases, if for no other reasons, but for me it’s because I know something of the simple human compulsion to just know.

Moreover, my own researches into company histories and the Rumaniacs Project showed that sometimes the bottlers themselves are no longer in business and so there’s nobody who can shed light on a bottle being queried; worse, in some cases existing companies themselves kept no records of what the hell they did. There was a Cadenhead rum bottled in 2003 which was practically unknown, to give one example; the SMWS’s lack of a list of the rums they themselves had issued was another, and I can assure you that almost no old rum-making company anywhere in the world has records of all its bottlings, blends, label changes or even marks – such things were either never deemed of great importance or simply left forgotten and unrecorded.

Books

There was a time less than a generation ago, when books were all we got, and we were grateful. Though not specifically created with the aim of compiling lists or catalogues except as an incidental by product of their researches, they immeasurably aided in such efforts. It was considered, with a kind of endearing innocence, a fairly easy task in the pre-Renaissance and pre-Internet era when most people knew at least something about Caribbean and Latin rums but rarely ventured further afield. Local rums in other lands and climes stayed local and developed their own national character, and at best it was world fairs and occasional newspaper articles over the last hundred years that allowed more knowledge to disseminate. Nobody ever really tried to collate or tie together the world of rum into a cohesive whole.

That said, these early books, when (or if) they made lists of rums at all, concentrated on geographical areas for the most part, and tried to gather some knowledge together with what limited information was then available. Excellent as they were in moving the subject of rums forward in the greater perception of the drinking world, however, they had several drawbacks.

For instance, the absence of reference or supplementary materials made it necessary for authors to do primary research, in person. Rum lacked the cachet of wine and whisky, where print magazines and newspapers had on-staff critics who were sent on the tab to major wine- and whisky-producing regions to taste, interview and record: in stark contrast, aspiring rum writers were a solitary bunch working in obscurity, and they had to travel and research and experience rums on their own dime. Unsurprisingly, therefore, they stayed within the confines of the regions with which they came from and which were accessible, primarily the Americas and the Caribbean; and they ignored the rich pickings to be held in other parts of the world (a weakness which continues to this day).

Moreover, by the time any book was written and then proofed, sent to printers and distributed, it was often already overtaken by new releases, and if not, became so within a year or two. Once printed they were locked, and so they dated fast. At the time, the majority of the rum market consisted of a sea of blends (only occasionally re-released, re-branded, or reformulated), and the era of multiple annual releases by a host of independent bottlers or multitudinous monthly batches by micro-distilleries, had yet to arrive: but, even with this slower pace of rum releases, no book could ever really stay current.

Ed Hamilton’s Rums of the Eastern Caribbean and Complete Guide to Rum from the 1990s — both based on his extensive travels and distillery visits in the region — have long since become almost obsolete (thought retain much usefulness as snapshots in time), and even a more recent book like Martin Cate’s Smuggler’s Cove has a rum list that is at best representative, and is approaching its sell-by date as new rums and distilleries emerge on the stage. Other recent books like the French language Le Guide Hachette des Rhums (The Hachette Rum Guide) with originally 400 and now 550 entries, or Alexandre Vingtier’s more modest effort 120 Rhums are useful additions, but unless updated, will suffer similar fates. And the multi-kilo double-tome of the recently printed Caroni distillery history and its bottlings will surely have to have a companion volume to account for all the releases that will be made after 2022.

The original Encyclopaedia Britannica tried to address the same issue by printing annual yearbooks where they updated the content as best they were able. But aside from the Hachette guide, the writers of books on rum never went that far (and let’s be honest, why should they?) – they rested on their laurels as published authors and moved on to other projects. Even something as potentially useful as a summary-form Rum Bible (an equivalent to Murray’s work on whisky) was never written, because nobody was in a position to taste the hundreds and thousands of rums such a book would entail, even assuming they were known or available for tasting. In any case, any rum lists included by the various established authors were seen as adjuncts or extensions to their main work of description, story telling and historical recollection — not the primary focus of the work itself.

Things started to change with the advent of the internet and the rise of enthusiast driven weblogs, which started around 2007. Most of the early efforts in this direction were rum reviews, and sites like Refined Vices, Rum Reviews, El Machete and others stuck with this formula until they went dark and were replaced by yet others doing the same thing. Websites were and almost always are, run by individuals, and such initial forays into the online world came from this pool of enthusiasts who did their best to create, as best they could, a repository of the rums they had tasted. Few went further, though some certainly did take it to the next level – and such sites often dispensed with the whole reviewing gig altogether, perhaps as they had to.

Ed Hamilton’s Ministry of Rum Website

Possibly the most influential of the early rum-focused websites, the Ministry of Rum was launched in 1995 at a time when usenet groups and dial-up bulletin boards dominated the online space and user interaction. Windows 95 debuted that year and the Netscape Navigator had only been released the year before and the internet was a wasteland of disparate websites only gradually finding their way. Easy-to-use website builders like WordPress and SquareSpace were far in the future and Ed hired a programmer to create his website. He added a discussion forum for users, wrote some commentaries, added some articles, but for the purpose of this essay, it was his brief country distillery listings and the rums these distilleries produced which is of note. When I researched the early reviews of my own, it was often the Ministry that provided the first core data points of a rum’s origin, stills, strength, company background and other products they made. Sadly the site is moribund and most of the links do longer work, and there was never any structured table listing one could consult, so I actually have no idea how many rums were under the hood.

The Burrs: Rob’s Rum Guide, Ultimate Rum Guide, et al

This is what led to one of the earliest websites that tried to capitalise on the burgeoning rum scene of the late 2000s and early 2010s: the Burrs’ Rob’s Rum site, and the associated list of rums which topped out at 622 items (but which lacked many of the minimum provided details we now take for granted). The site was part of an overall multi-channel effort that tied into their various commercial enterprises, especially the Miami Rum Renaissance (which at one time was the premiere North American rum event). Their Ultimate Rum Guide (now offline, and ported to Instagram) was another offshoot of this approach and listed some rums and provided brief details. Unfortunately it was never scaled up or maintained, ignored far too many rums, was limited in geography, and I don’t think it’s been updated in a while. The efforts of the Burrs have been redirected to the American Rum Index, the Rum Minute (60-second tasting notes on You Tube) and other businesses in which they are involved. So the whole database “project” (and it never really was anything so structured or grandiose) really didn’t go anywhere and died for lack of oxygen.

Taster’s Guide

Another attempt which was much more serious was the (now defunct) site of Taster’s Guide, created around 2010 by a longtime correspondent of mine named Stefan Hartvigson from Sweden. Over time it amassed what for the time was an enormous listing of popular rums – it’s now dark so I can’t remember how many rums it ever had, but it was very detailed, and had many of the fields enthusiasts were to clamour for as the bare minimum in years to come: name, age, components, source, country, distillery, strength and if available, year of distillation, plus notes on each distillery and other rums they made. The site never got the acclaim it deserved because Stefan – a marine engineer by trade – never marketed it with that intent or did more than casually update it — like many such sites (including my own) the initial impetus for its creation was simply to catalogue his own purchases and info he picked up along the way. Gradually this grew legs and he tried to keep it going with an enormous body of research, but by 2015 he acknowledged that it was too much work for one person to do, and he let it go.

Peter’s Rum Labels

A site that defies easy categorization and is not a database in the strict sense of the word, but was and remains enormously useful and probably one of the best out there for what it is, is the Czech site of Peter’s Rum Labels, created and maintained in English by Petr Hlousek from Prague. It does not have a standardised database format, and doesn’t try listing anything. What it has is pictures of rum bottle labels, and data on each company that makes them, plus translations and “the fine print” on each label. This might not sound like much, but from a historical perspective its worth is incalculable because of the 9,785 label pictures he has from nearly 6,000 producers, companies and brands, many predate the modern era and provide a window on rums of years and decades past. Moreover, there are often small company bios accompanying each (the site is more or less organised by countries and producers) and even how many medals a company or its products won (though I think this ceased around 2010).

Rum Ratings

Then there was Rum Ratings, a website initially created to be a repository of tasting notes by Andrew Shannon, which went live in 2012. As a student in the UK he wanted to remember and catalogue the collection of rums he had left behind in the US as well as those he wanted to try in the future, and the site began life as a personal blog in which he kept his own scores. As he recounts, “Within days of launching, people somehow found the site and asked me if they could enter their ratings as well. It took a little work, but after I opened it up to others things just seemed to take off.” Even without any sort of deliberate or conscious marketing the site gained popularity, perhaps because it was the only one of its kind in the world – a place where people could fulfil their desire to record their own scores and comments of rums they had tried.

The site has come in for criticism (including by me on occasion), because of its populist ethos, something Andrew is correcting over time by bringing in links from external bloggers. The average scoring method is problematic when there are only a few ratings (it comes into its own with greater volumes), though the bar chart of score-distribution is great. The data set for each rum is also somewhat limited and as a rum lover I confess to always wanting more.

Yet I’ve come around to really appreciating this site — because alone among all the others it does get updated, you can post your own comments, and you can rate a rum, of which there are close to 8,600 as of 2022. Moreover, because it has been around for so long, it has opinions on rums that go back a long way, which is a useful window into the past (I made use of that when demonstrating why the Diplomatico Reserva Exclusiva had to be considered a key rum, for example). So as a resource of archival material, it’s really very useful and should not be discounted simply because it is user-driven and lacks rigorous reviews or tasting notes. It remains in use, updated and non-monetized by a person with no development experience and no connection with the rum world at all and I continue to have a real affection for Andrew’s work, and use it regularly. (Note: there’s an app for it now, but I prefer the desktop version).

Reference Rhum

When it comes to pure data shorn of any externalities, perhaps the best pre-app, pre-mobile, pre-wiki website database of rums, which has now been offline for many years, was the French site of Reference Rhum (not to be confused with the sales page of that name which now exists, or the review site Preference Rhum). Reference Rhum was the last gasp of websites curated by a single person which sought to list all rums in existence, and honestly, I still think it did a bang-up job and came as close as anyone possibly could at that time, to nailing it. There were tons of stats for each rum, label or bottle photographs, distillery notes, distillation notes, proof, age, dates, names, sources. When I was doing bottle lists for some of the “Makers series” company bios, it was Reference Rhum I went to as my first stop. At its peak it had around 9000 rums or more listed with a level of detail no other site even came close to, except for Rum Ratings (and the two weren’t comparable). Sadly, the job of updating and curating the site became so onerous and time consuming, that the owner finally shuttered it, which was a loss to the rum community that is often not appreciated.

The position in 2018

By 2018 or so, whether acknowledged or not, it was clear to many that there was a huge gap in the reference materials available to rum aficionados globally with respect to actual bottlings. Books that were published and posts that were put up were all about rum companies and production details, historical perspectives and limited or specialised foci; and we were and are immeasurably enriched by the research of Matt Pietrek and the few others like Marco Pinieri, Anil Lutchman et al, who mine this lode. Yet aside from the imperfect examples listed above, no-one has ever tried to list all rums in production or those from the past, perhaps because the job is just so absolutely Himalayan in scope. And after my near complete failure to find any reference to that Cadenhead rum mentioned above even from the bottler, I began to realise this could turn out to be a very serious issue indeed for future buyers, fans, writers or researchers.

However, even as the internet widened and democratised the expertise of rum pundits (and their number), it became equally obvious that it was almost impossible for any single individual to create, curate and maintain a master database of this kind. Given the volume of rums and brands available around the world, and adding to that the historical one-offs, merchant bottlings, independent bottlers or special editions dating back (in some cases) centuries, it was simply too time consuming. It would require full-time effort, not occasional after-hours dabbling by enthusiastic amateurs. Nobody has that kind of time in our world, quite simply because nobody is getting paid to do it and it’s such a thankless job. A new system of such record keeping therefore had to be found to address the lack of any serious databases of rums in existence and the gradual move away from desktop computers or even laptops.

We’ll discuss the one application that tries to crack this issue, in more depth in Part II

Other Notes

  • The site of Spirit Radar is an interesting one. Registered in 2020 and run by a small team out of the Czech Republic, the site notes that it is a “next generation data platform for rum and whisky collectors.” They monitor auctions, online shops and ecommerce sites for historical and current bottle pricing information for rums and whiskies (some 60,000+, they note). The site is fully commercial – you pay for the data service and pricing information and have the option to do a free 14 day trial. As part of the data on each bottle, the sort of thing we need — country, strength, age, distillery and so on — is included, and there are options to create a bottle list of your collection, and the site shows its aggregate current value. Because of its collector and commercial focus and inaccessibility to the broad mass of users, I elected to not include it – but it is a resource of the kind this article speaks about.