BeerMapping.com Map Critique
For my critique, I decided to go “off-map” with a project that much informs my own and contains valuable lessons about the organization of data in a project that bears a similar entry-structure to our own. The Beer Mapping Project is a user-driven interactive map based on the Google APIs that charts the locations of breweries, brewpubs, beer bars, beer stores, and homebrew stores in the United States and around the world. The Beer Mapping Project launched on October 31st, 2005. In 2007, it expanded with a forum and a built-in reviewing system for users. As of this writing, there are 13,216 locations listed on the map worldwide, with around 75% of these being in the United States. To give you an idea of activity, around 160 total map entries have been made in the last month, zero images have been added, and 47 reviews have been created from October 5th to November 5th. Going forward, I will talk about how the map looks and what interface adjustments I would make.
The Beer Mapping Project (BMP) is best approached as a pure data-modeling map. From the Elena Pierazzo quote that Rory mentioned in a prior reading: “[Data]-Modeling can be seen as the intellectual activity which lies at the base of any computational effort, namely the methods and the languages we invent to communicate our understanding of a particular cultural object.” For this particular effort, BMP’s data is given structure from the Google API, creating pins that contain rich content regarding the location. Given that the data BMP users have collected can be repurposed (and has been) for other displays, I will focus on how the map displays the data and some of the potential shortfalls.
Locations are split into two parts. The Regional Maps legend contains only Breweries and Brewpubs, while City Maps legend contain these plus Beer Bars, Beer Stores, and Homebrew Stores. City Maps are loosely centered around the city’s center and extend into the surrounding metropolitan area. BMP lists a total of 43 cities worldwide(with a very US-centric basis). Regional Maps are broken into countries, 12 international (the international maps usually contain the city legend items as well), and the United States. The US is divided into six regions total to accommodate for the high volume of postings. In the maps you still have full range of motion around the region (or world), however only data plots appear for that region.
The need for both maps becomes apparent around major metropolitan areas. Zooming in or out of the map does not alter the amount of pins on the map. Users can hover over the pins to reveal the name of the brewery (and any reviews). Upon clicking the pin, the user is directed to a sign that provides things such as name, address, website, directions, score, and links to other beer-related websites that can provide insight on this listing (They are not directly tied in via any API, rather they are using Google Boolean logic to achieve the desired point).
Upon clicking the link within the pin’s signage, you are brought to the page that shows the name and address, a close-up map of the pin’s surroundings, and if added by users, photos, reviews, and amenities.
While the map is authoritative in the data its users have collected over the last eight years, there are still serious flaws in how some of the data is presented and collected. Much of this has to do with the bias brought on of its own user-base, and horizontally-inclusive datapoints that lack expansion.
For all items tagged as “beer bars,” there is no hard-line criteria given by the moderators of the map beyond “use your best judgement.” As this could be any bar or restaurant, the difference between one datapoint and another can be vast. A cursory look at the data points for beer bars correspond significantly with high-income and high-traffic areas with younger populations. This leads me to wonder if some areas don’t have a place worth sharing, or the neighborhood is unlikely to overlap the demographics of the userbase, or both for the same reason.
The maps search and listing features leave much to be desired as well. On the map page, a full list of the breweries contained within is in alphabetical order. There is no means to search through this list, the search tool is not integrated within the map itself and instead goes through the entry pages. This entire list does not indicate what these entries are, it only supplies names. Next to some names are numerical scores from 1-100, indicating the location’s review.
On the entry pages themselves, much of the information is redundant or incomplete. While there is the rather nifty ability to redraw a map in a defined radius around a certain point, there’s no way to refine the results to include only one type of listing. Most breweries have an empty photo section, and no reviews. Those that do, usually only have one or two, unless the brewery is considered “nationally established.” These pages don’t take into account differences between types of listings, so no dynamic information can appear that might be more useful for one link over another.
The link submission process may stymie the addition of rich content to the site. If a user wishes to submit a site, he must wait for a lone moderator to verify its existence. In this time, the user cannot post a review, or add photographs, and the timeframe is murky as to when this information can be verified (I submitted a location recently out of curiosity for timeframe, will report back on how long it ended up taking). Furthermore, there is no means of corrections or deletions without direct intervention from concerned users. While the moderators are quick to rectify issues or postings brought to their attention, they are not actively perusing their 13,000+ links for closings, name changes, or information updates.
In my proposed redesigns of this map, I want to focus on what makes BMP great, and that it is, quite simply, the most comprehensive interactive map of beer related-locations in the world. Where it fails is organizing its information for users to successfully peruse, and attaching ancillary and underdeveloped databases onto its location data. What I propose is for BMP to integrate itself with other already successful communities.
First is adding value to the entry pages. Currently, their in-house review system is being held up by a small but dedicated group of users. With an unfocused rating system and the content buried within forum privileges, it would be best to pull the database from a more established site such as ratebeer.com or beeradvocate.com. This will give users a better sense of what the brewery has to offer.
Brewery/Brewpub
Entries for restaurants and stores should have data that is similarly tailored to their offerings. Thankfully, there are other databases, such as BeerMenus.com, that provide an up-to-date listing of the beers available at high-end stores and specialty bars.
Specialty mockup: http://i.imgur.com/STjH1b9.png
With this added data, there are reliable cross-referencing datasets that can be used to refine the map. Below is my final mockup. Certain legend points can be engaged or disengaged. Beers can be searched for in style and name. Results can be filtered by ratings. I see some limitations in the ability for cross-referencing data against external databases (for example, applying a ratings filter that pulls the ratings from one site to check the beer selection list originating from another site to create a map of restaurants with the best beer), however the data is now much more navigable overall.
Mockups
Names
Styles
Ratings
While some mobile apps have taken advantage of the BMP’s data to inform users of their proximity to map points, much of the data rests in what now feels like a very old incarnation of the interactive web. This map and its workflow for submission were clearly built in the pre-smartphone era, and I believe that if this map were recreated from scratch today that entries would likely originate from a smartphone app rather than a submission form on a forum waiting to be approved by a moderator.