Notes on an Epic Night redesign
Epic Night was a nightlife-stories-and-venues site Alyssa was building in 2009. After receiving a long redesign proposal from a designer collaborator, she walked through it line by line. The original proposal isn’t preserved; what follows is her side, which reads as a fairly complete catalog of the design tradeoffs she was thinking about at the time.
Thanks — lots of good stuff in here.
The ratings system you’ve outlined seems good in principle, but too complicated. Our number-one goal should be to make the site easy to use, and KISS is a large part of that.
If the story is long, it just pushes the comment/ratings widgets down to where the story ends — on the theory that users will want to rate or comment on a story right after they’ve finished reading it.
I agree the search box shouldn’t be so prominent (e.g., Reddit’s doesn’t see much use). Do you have an idea for what you want to replace it with?
We should put something saying “Interested in advertising on our site? Contact us at blah@epicnight.com” on the Academy page, so we can put banner ads up when the site gets popular enough.
The half-life I’ve been using for stories is about five hours, so that for every five hours that passes, twice as many people need to rate the story positively for it to remain equally high on the “Top” rankings (which determine the featured story). That seems like a good balance between stability and putting up fresh material.
Putting “Write story” before “Stories” is a great idea. The site is going to live or die on whether people are writing lots of stories, and we want to encourage that.
There’s only one line of whitespace separating story text on each side, and I think it would be really awkward to get rid of it. We want users to be able to visually identify the text very easily so they start reading it, not have it lost in a jumble of lines.
Including a venue’s rating on the map is a good idea and pretty easy to do.
A five-mile radius doesn’t work very well, because the number of venues within that radius varies hugely — in San Francisco it would be the entire city with a million people, while in the town I went to college in last year it would be just 30,000 or so people. I’ve included all venues in the same zip code, which matches population density pretty closely.
Star ratings, although they look cool, are pretty tricky to implement technically, and so they should be done only if there’s a few days of spare time available.
Adding lines of story text to the venue page might be a good idea but would break the table format; I’ll have to see how it goes.
I just put in venue statistics on the venue pages yesterday (e.g.
epicnight.com/venues/11 ).
It would be a lot of work for us to research and then manually update bar specials for a hundred venues all over the country, let alone a thousand, let alone ten thousand (which we will get to eventually if the site becomes reasonably popular). As I’m sure you know, running a startup is a ton of work all by itself, and I don’t think you’ll have the time.
There’s no easy way I know of to record page views, which is why, e.g., StatCounter is so popular — even though it charges you if you want to log more than five hundred IP addresses.
Including number of comments with a user’s data is a good idea and pretty easy to do.
I think publicly displaying author location is a VERY bad idea. Lots of people are going to be posting things on this site that they wouldn’t want their boss, their wife and kids, their church friends, etc., to find out about.
Sorting stories and venues by various parameters using AJAX would be tremendously useful. I think that’s what I’m going to spend this weekend on.