Barley Mowat 

Archive for the ‘StAugustinesVan’ tag

High Tech Meets Beer

with 2 comments

I was over at St Augustine’s the other day and noticed that one of the TVs had a weird display of symbols instead of a sporting event. It looked something like this:


Huh. I was right.
The real world WAS blurry that night.

That, in case you can’t quite make it out, is a display of all the beers currently on tap, and what percentage of each is left. This is a very encouraging development for a few reasons. First, the days of making a drink list in your head and then progressing through it only to find out that James’ One-of-a-kind IPA What He Brewed When He Was 15 But Lost Then Refound is now a dead keg are now gone. Keg is low? Order it first.

Second, the video list is always perfectly up to date. Printed menus tend to become obsolete before they come back from the printer (or photocopier in Nigel’s case).

That’s about it, for now. There are a bunch of enticing possibilities that adoption of this tech allows. Once you have this information in a format that lets you display it here with a fancy graphical user interface, it means you have it in a database somewhere (unless you paid your software guy with beer which, let’s face it, is likely what happened here). Digital tap lists let you do a few cool things if you’d like:

  • Update your website tap list in real time.
  • Tweet new kegs as they are tapped automatically. Likewise tweet warnings at the X% mark before you run out. (Should likely create a new Twitter account for these to avoid drowning out non-beer tapping news)
  • Breweries can show you where to find their product, in real time.
  • How about an iPhone app (or web interface) that lets you pro-actively order your beer, or set a drink list in advance?
  • You can calculate most popular or least popular beers dynamically, and share that info.
  • Display Ratebeer rating, or Twitter ranking.

Yeah, I’m a bit of a software nerd (it’s what I do professionally… what? You think this blog makes me any money?). I’d be very curious to talk to whomever wrote this bit of code for St Augustine’s, as I once designed a very very similar system unbidden for the Alibi Room with hopes of convincing Nigel to give me free beer in exchange for it. The trick is how does one actually calculate the percentages? It’s not as easy as you’d think, since there are precious few good ways to figure out what’s left in a keg (or to account for different keg sizes). Here are the options:

  • Stick each keg on a digital scale and weigh the fucker. Current weight minus dry weight equals beer weight. Pros: Accurate. Cons: Crazy expensive.
  • Slap a digital flow meter in each line. Same pros/cons as above.
  • Note that there are approximately 100 16oz pints in a 50 litre keg, and subtract 1% for each pint pulled. Pros: Fairly accurate. Cons: Requires either perfect integration with your POS system or religious adherence to a pint-tracking app by your bar staff.
  • Do the above, but also determine a baseline time that each keg stays alive for, then predict the end based on reported pints, average lifetime, and good old fashioned statistics. Pros: Most accurate, adaptive to lazy staff. Cons: Statistics are essentially magic and hence the devil’s work.
  • Good, old-fashioned, dead reckoning, updated periodically by hand. Pros: Cheap and easy. Cons: Kind of defeats the purpose.

I figure they went for one of the last two, but who knows. Can someone from SA’s enlighten me?

Lastly, the display above, while appreciated, definitely needs a few tweaks. So if anyone is listening and/or taking notes, here’s my quick-hit gripe list. You knew it was coming, didn’t you? Well, prepare to be surprised. Since I do software professionally, this complaint list is a bit more put together than my normal stream of profanity-laced sub-coherency.

  • The beer pint pie chart, while very cute and likely smart looking on the desktop monitor you designed it on, is virtually unreadable and nigh-useless on the big screen. What’s the difference between 50% left and 25% left? Yeah, I can’t tell either. All it does is take up space, which leads me to…
  • The per-beer widget is too big. Too much space is spent on design, which would be okay if it was visually appealing. Look at this display. The bits of information you need are: Brewery, Beer, Remaining. Everything else is window dressing and can be cut.
  • Grids don’t work for presenting lots of information. Humans like to read lists, which means moving your eyes top to bottom. Use two columns of wide, short widgets instead. You’ll be able to fit many more items on the page AND it will be much more legible.
  • Again, the small number of items per page means lots of pages. This means cycling through the pages quickly and often to make sure all the information is available. That means customers will miss what they’re looking for, have to wait through all the other pages, and try again to find it. This can cause frustration and a tendency to give up and return to the printed menu after investing a few fruitless minutes. These folk are now slightly pissed off, and they haven’t even had a beer yet.
  • How are the beers organized? I’m going to bet by tap number, right? You know what’s useless to clients? Yeah, that. How about organizing them by colour. You have the info (note the pint glass).
  • How many other kegs are left? If we run out of house lager, I’m betting there’s a fresh one right next to it, but other beers might not have backups.

All in all, folks, it’s a first version so we need to give them time and patience. Feedback doesn’t hurt, though. Just remember how things were a few weeks ago before this system was rolled out. In a year or two we’ll wonder how those luddites of the dark ages even managed to order beer at all.

Written by chuck

June 30th, 2011 at 10:35 am

Posted in Beer and You

Tagged with