Disagree. It's the whole point of the page, so it isn't excessively redundant if you say it more than one way... I'd just put a single column that's a different color (ie, green) for the best day, maybe with an embedded checkmark or some other indicator for the color blind.
1. The URL - would be better if it's a dedicated domain name (e.g. whenshouldigo.com).
2. I agree with the parent about the color coding. It might be more clear if you use a red-green spectrum for each day (e.g. red for the evening, when things might be more busy, and green for the morning, when things might be less busy).
Very nice job. I don't have a use for it since I'm not in London, but I like your implementation and design. A comparison feature would be nice, so if you're uncertain where you want to go one day, you can compare activity.
This is pretty cool. If I were in London, I'd use it.
You could integrate Google Maps to show you the least busy places in a certain part of London, so it could be used for "I've got an afternoon to kill. I'm over here; where can I go?"
Yeah, make it a smart phone app, charge $2.99 for it and include everywhere that foursquare supports, money to be made! Although maybe foursquare could just release a similar feature.
One immediately apparent reason would be API rate limits from Foursquare. It's easy/practical to query for "who's here" for a limited number of venues a few times an hour, start doing it for hundreds or thousands of venues and you'll start running into problems...
Yes it could be expanded. Because the Foursquare api doesn't provide historical data it takes a few weeks of monitoring the venue before I have useful data for it however
Cool idea - is the % of foursquare users in a sample population consistent? What if Tuesday is "bus in the foreigners/elderly?" - valuable info regardless, but I'd be curious to know how that holds up.
It's always interesting introducing a feedback loop into a complex adaptive system. I wonder how close to an even distribution this would make attendance if it became hugely popular.
Nice one. This will help me avoid the sort of people who use Foursquare, or who like playing with their mobile devices while in art galleries. And, I'll assume, will help ensure I have plenty of opportunity to meet others like me.
Another cool feature would be a temporal zoom - let me see that data juxtaposed over different months or weeks (I imagine it likely that the busy/free days vary depending on the season); conversely, it would be useful to be able to zoom in to the hour for specific days (best hours to visit today). That would help either plan vacation visit agendas, or determine an intra-day itinerary.
I liked the idea. It could have been more interactive, may be using Flex charts or any thing else for that matter to give cool effects like... umm wait.. what are the units on the Axes ? Give a number on the bar when the user hovers the mouse over them.. Its nice work for a 24Hr time line though...
I like your idea but what are you going to do when a lot of people start visiting on the least busy days turning them into busy days? I know from personal experience that Wednesday afternoons between 1-4pm are usually the quietest times in movie theatres and other venues that tend to be packed on weekends so by putting this information in people's hands you're messing with my zen dude.
Wednesdays used to be quietest for cinemas in the UK but now it's one of the busiest thanks to the Orange 241 offer that's been running the past few years
But since this system only incorporates data from the past few weeks the shift towards quite days will be detected pretty fast and thus these days will start to be visible as busy. It will balance itself out I think
Since only a small percentage of the UK population use Foursquare I'd be more inclined to be concerned it might head in the opposite direction (i.e. relatively small numbers of people using this site to identify quiet days to visit are substantially more likely to check in - resulting in more noise than signal)
Another reason software is a bottomless pit of opportunity. There are so many things that can be optimized if only the data can be put into the right hands in a usable form such as you've done here.
Looks great, definitely a value-add yet very minimal. I sent my friends in London to it, could see it going a bit viral.
I found venue navigation very confusing. When clicking on the underlined 'tate modern' (top left) I expected to be given a choice of other venues. How do I see a list of venues? I can only discover them via the 'top lists' and 'Random'?
Thanks. I certainly need to add some kind of index/homepage that lists all the venues. The current 'venue roulette' navigation isn't exactly the best approach.
Good call on using foursquare to do the work. I'm not sure if you can use this source to include the best times of the year also, but this would be ideal. When people ask themselves (or others) "when should I visit?" they are often referring to "what time of the year should I visit?"
For a 24 hour job, I'm wowed. I can see a lot of expansion potential, notably in adding location (by country, state, city) and categories (i.e. breaking it down by museums, theaters, parks, etc.).
Consider making the graphs into a widget for local tourist, hotel, and city pages.
Whoops. I should integrate with http://opening-times.co.uk/ perhaps. Although places like the fantastic Soane and Geffrye seem to get very few Foursquare checkins even at the busiest of times
opening-times.co.uk doesn't looks like it carries museum opening hours. Actually a semweb data source of visitor attractions and their opening hours would be a very valuable resource. I don't think one exists.
you're not measuring how busy a place is, but rather matching demographics between foursquare users and the various museums.
I wouldn't be surprised to find that foursquare check-in's is a very poor proxy for this kind of venue. For bars/clubs and coffee shops maybe, but you're measuring with the wrong stick. Having said that, it does seem to confirm common sense (that most places are busiest on the weekends and mid-week is quiet).
Shouldn't you switch the columns with museums list to see which one is least busy first? This will be more consistent with your supposed goal and will correspond with the fact that you say which museum is quieter first above (just below the bars).
edit: also, why can't I search or see a full list of all places?
edit: Also for SEO you should change the URL to something keyword-packed like whenshouldivisit.iamdanw.com/when-is-british-museum-busy and offer alias short links like whenshouldivisit.iamdanw.com/british-museum, etc.
Glass-half-empty vs graph-half-full problems abound in UX and the answer is always to use the one that tells the least lies. So in this case the graphs should represent people - how can you represent emptiness?
I would like to see actual people on the graph (not one person scaling) as this will show you exactly how busy each place is - and, if you know how big a place is you can estimate how full it will feel.
> actual people on the graph (not one person scaling)
yeah I couldn't find an image which showed exactly what I meant, but yes, that is what I had in mind.
Although you seem to be suggesting that you should be able to get a literal exact count of the people. I think that that won't be possible or useful because the data is aggregated over several timepoints. I guess "on mondays, approx 10 people visit" might be useful, but I see this graph as more of a general indicator of how busy a place generally is.
Using a repeating image of little stick people will hint to users that the bars represent visitor volume.
>Using a repeating image of little stick people will hint to users that the bars represent visitor volume.
This is a very good idea. Unfortunately his bars are vectors drawn with SVG, I don't think it allows background images.
Another, perhaps not as clear suggestion would be to simply add a "Smaller bars mean less busy" message under the graph, like you see on performance benchmarks.
It would be great to see some real numbers on these graphs as there is a danger that some days are being reported as really busy when the n is really low.
I made something somewhat similar a few years back because I thought it was something that I would use. I did not follow through, but I'm glad you did. I think there's a lot to be done with this. Good luck!
Great project. Is it possible to include the best time of the day to visit, e.g. "quiet on Wednesday mornings"? Also: Do you adjust for longer opening times? Maybe some venues close sooner on some days.
60 comments
[ 2.8 ms ] story [ 112 ms ] threadGoogle "comparison chart" and see how <a href="http://www.google.com/imgres?imgurl=http://printgreener.com/... facets</a> can inform a viewer in a simple way.
2. I agree with the parent about the color coding. It might be more clear if you use a red-green spectrum for each day (e.g. red for the evening, when things might be more busy, and green for the morning, when things might be less busy).
Here's my suggestion: ditch the black borders on the columns and add very subtle shadows.
I really dig the idea, by the way.
You might also want to take a look into incorporating Spotrank data (http://www.skyhookwireless.com/spotrank/examples.php).
SimpleGeo is one way to do that (http://simplegeo.com/).
You could integrate Google Maps to show you the least busy places in a certain part of London, so it could be used for "I've got an afternoon to kill. I'm over here; where can I go?"
referenced stylesheet "all.css" returns 404
The raphael code could do with cleaning up too
Another cool feature would be a temporal zoom - let me see that data juxtaposed over different months or weeks (I imagine it likely that the busy/free days vary depending on the season); conversely, it would be useful to be able to zoom in to the hour for specific days (best hours to visit today). That would help either plan vacation visit agendas, or determine an intra-day itinerary.
Another reason software is a bottomless pit of opportunity. There are so many things that can be optimized if only the data can be put into the right hands in a usable form such as you've done here.
I found venue navigation very confusing. When clicking on the underlined 'tate modern' (top left) I expected to be given a choice of other venues. How do I see a list of venues? I can only discover them via the 'top lists' and 'Random'?
Well done.
For a 24 hour job, I'm wowed. I can see a lot of expansion potential, notably in adding location (by country, state, city) and categories (i.e. breaking it down by museums, theaters, parks, etc.).
Consider making the graphs into a widget for local tourist, hotel, and city pages.
Spotted that immediately, but smart work anyway.
opening-times.co.uk doesn't looks like it carries museum opening hours. Actually a semweb data source of visitor attractions and their opening hours would be a very valuable resource. I don't think one exists.
I wouldn't be surprised to find that foursquare check-in's is a very poor proxy for this kind of venue. For bars/clubs and coffee shops maybe, but you're measuring with the wrong stick. Having said that, it does seem to confirm common sense (that most places are busiest on the weekends and mid-week is quiet).
Shouldn't you switch the columns with museums list to see which one is least busy first? This will be more consistent with your supposed goal and will correspond with the fact that you say which museum is quieter first above (just below the bars).
Cool project!
At least that's assuming that you'll want to visit at quiet periods. If you expanded to include pubs and clubs, popularity might be more desirable.
Perhaps change the graphs instead of being block colour to have a texture of little people - eg something like http://static2.bigstockphoto.com/thumbs/9/8/7/large2/789645.... or http://www.vectorstock.com/composite/49471/statistics-graph-...
edit: also, why can't I search or see a full list of all places?
edit: Also for SEO you should change the URL to something keyword-packed like whenshouldivisit.iamdanw.com/when-is-british-museum-busy and offer alias short links like whenshouldivisit.iamdanw.com/british-museum, etc.
I would like to see actual people on the graph (not one person scaling) as this will show you exactly how busy each place is - and, if you know how big a place is you can estimate how full it will feel.
yeah I couldn't find an image which showed exactly what I meant, but yes, that is what I had in mind.
Although you seem to be suggesting that you should be able to get a literal exact count of the people. I think that that won't be possible or useful because the data is aggregated over several timepoints. I guess "on mondays, approx 10 people visit" might be useful, but I see this graph as more of a general indicator of how busy a place generally is.
Using a repeating image of little stick people will hint to users that the bars represent visitor volume.
This is a very good idea. Unfortunately his bars are vectors drawn with SVG, I don't think it allows background images.
Another, perhaps not as clear suggestion would be to simply add a "Smaller bars mean less busy" message under the graph, like you see on performance benchmarks.
e.g. http://whenshouldivisit.iamdanw.com/venues/2661544
Is this just one person each day?